Search tips
Search criteria 


Logo of eurojageEuropean Journal of Ageing
Eur J Ageing. 2007 June; 4(2): 59–69.
Published online 2007 May 16. doi:  10.1007/s10433-007-0047-3
PMCID: PMC5546252

Exit and voice: an investigation of care service users in Austria, Belgium, Italy, and Northern Ireland


The past decades have seen an introduction of market elements in the provision of social care services (Finer 1999; Mabbett and Bolderson 1999). Welfare state reforms all over Europe have produced welfare pluralism and claims that the increased choice will enhance user participation, promote older persons’ autonomy, and improve the quality of services. Within the Fifth FP Research Project CARMA (Care for the Aged at Risk of Marginalization) a case study among users of care services in Austria, Belgium, Italy, and Northern Ireland was conducted that focussed on friction and conflict between clients and service providers and investigated the reasons for discharge and denial of admission to a service. The data from this study can be interpreted in terms of Hirschman’s (Exit, voice, and loyalty: responses to decline in firms, organizations, and states. Harvard University Press, Cambridge, 1970) theory on ‘exit’ and ‘voice’ as expressions of consumers’ dissatisfaction with the quality of a product. Data were collected in different systems offering a variety of procedures for exit from one provider and the choice of a competitor. Also different practices of handling voice i.e., complaints have been documented. The paper questions to what extent various possibilities for exit and voice can enhance users’ autonomy and increase the quality of the service supply. It thus contributes empirical findings to a debate that often emphasizes ideological arguments.

Keywords: Care services, Comparison, Exit and voice


This paper contributes to the debate about the effects of introducing choice and market mechanisms into the provision of social services for older adults (Lundsgaard 2006; Rostgaard 2006; Kremer 2006; Vabø 2006; Clarke 2006; Baldock 2003; Ungerson 2003, 2004; Sipilä et al. 2003; Tester 1996). A mixed welfare economy (Evers 1991) is claimed to empower care dependent service users by transforming them into customers who can choose between competing public, private non-profit and private for-profit care providers. Furthermore, cash allowances paid directly to care dependents are said to enhance their purchasing power to pay for informal care and hence ‘commodify’ care (Ungerson 1997). While some observers welcome the consumerist approach as a means to “strengthen the position of the older people as active consumers, making their individual demands clearer” (Lundsgaard 2006, 366), others warn that “[c]hoice mechanisms carry the risk of reproducing economic inequality” (Clarke 2006, 436).

The present paper investigates the relationships between service users and service personnel through a qualitative case study of services in Austria, Belgium, Italy and Northern Ireland. It gives an insight into the range of choice and the extent of service users’ power in four different variations of welfare economy. While Austria and Italy offer cash allowances and a moderate (Austria) to low (Italy) level of formal service mostly provided by local monopolies of non-profit organisations, the Belgian (to be more precise the Flemish) system has comparably extensive services provided by competing non-profit and public agencies. In Northern Ireland the users in the study received care from the free public services of the UK National Health Service.

Our particular focus was on so-called deviant users, whom staff regard as ‘difficult’ to provide care for. The objective of studying ‘deviant’ cases of users was to look for any implicit preconditions a client had to meet in order to get adequate service and support. Given that these conditions for care use can be regarded as operationalizations of the concept of social exclusion we extended the approach of Blackman et al. using social exclusion as an “evaluative concept across different welfare cultures” (Blackman et al. 2001, 4).1 The service units we studied represented differing welfare state offers to support and integrate older care dependent people: various forms of domiciliary care and day care centres. The services can be regarded as comparable given that they represent responses of the welfare state to prevent older care dependent people from becoming (increasingly) marginalized. The services in the study conform to Laing’s definition of long-term care, in that they provide “continuing personal or nursing care and associated domestic services for people who are unable to look after themselves without some degree of support, whether provided in their own homes, at a day centre” or in a care home (Laing cit. in Tester 1999, 136; cf. also Finer 1999; Mabbett and Bolderson 1999; Blackman et al. 2001; Anttonen et al. 2003; CARMA 2005). Accordingly we studied how and why excluding processes may affect those older persons that were identified by a welfare system as being vulnerable—i.e., who have bad health and/or lack access to e.g., economic, social, and/or cultural resources. Economic, social and/or cultural resources being denied to persons at risk of social exclusion—as conceptualised for instance by Kronauer (1997, 2002)—links the specific case of social exclusion in old age to the general scholarly debate on the concept (Rake 1999; Walker 1980, 1981).

We asked care service personnel to record and report incidents of friction, discharge or rejection of users on simple data forms similar to the complaint or error forms of quality assurance systems (see example in “Appendix”). The findings were extended by observations of the daily practice of service provision, as well as by interviews in units other than those in the main study.

The analysis of the data generated a theoretical model of care provision based on the practice of care for older adults in differing welfare systems. This model challenges the claims of market ideology that sovereign consumers direct services. Given that our theoretical model of care service provision describes relations of clients/consumers with service providers in a welfare market, a close look can reveal how much (or little) power the individual client actually has on this market. Does the reality of welfare pluralism live up to its anticipated merits? Does the care dependent person perceive the pluralist offer of services as a free choice? Does the option of choice enhance autonomy and agency? Can the service consumer exert power by her/his complaints or by threatening to choose a competing provider?

Neo-liberal economists have understood the free choice of the individual customer as the propelling force that leads to the most effective method of production, the highest quality and the lowest prices. This desirable result is attributed to an ‘invisible hand’ (Adam Smith) which aggregates the single consumers’ decisions in a way which is beneficial for all—the consumer as well as the society as a whole. The assumption of the beneficial effects of competition has become increasingly influential with decision makers (Folbre 2001, 48 ff).

According to free market theory, free choice will allow customers to abandon—i.e., ‘exit’—a company, which produces goods of unsatisfactory quality, and this exit will cause financial losses for that company. These losses in turn will motivate the managers of the company to act, or else the company will go bankrupt and disappear from the market. The free market model regards such an elimination as an acceptable loss, as it is a consequence of low efficiency, and on the whole the quality of the goods is preserved or increased.

However, in 1970 the economist Albert O. Hirschman introduced a novel way of analysing customers’ reactions to a decline in the quality of the goods or services they purchase on the free market. Hirschman’s point was that the free choice of the customer to choose another provider/supplier is but one reaction to a deterioration of the quality of the goods and services and is far from being a direct response to the company (Hirschman 1974, 14).

Voice’ is another channel for customers or citizens to express their dissatisfaction about deteriorating quality of goods and services to the responsible parties directly. Voice can take the form of complaints, petitions and protests, in short as organised expressions of dissatisfaction. In this study we interpret the behaviour of so-called ‘difficult’ clients as consumer dissatisfaction and hence ‘voice’.


Research was carried out within a multiple case study following the principle of triangulation in data sources, evaluators, perspectives and methods (Denzin 1970; Yin 2003, 98f). Similar to the trigonometric method used in surveying, a social phenomenon is thus studied from more than one angle. We collected data about friction, discharge and rejected requests for admission in a total of nine studied units in Austria, Belgium, Italy and Northern Ireland over a period of 12 months (from January to December 2004) by involving service personnel in data collection. A studied unit consisted of one provider organisation or one service unit of one provider. The services were non-profit home nursing or home care services for older people and two day centres for seniors.2 The local research teams were advised that the organisations they chose to study would be co-operative and willing to collect data over a 12-month period without any material reimbursement for their efforts. Thus, we focussed on this relatively small number of units for the study to ensure the quality of the data. Data sheet forms—as a standardised measure to collect data—were developed by the author’s research team, discussed with the local researchers and revised according to their suggestions (see example in “Appendix”). Hypothetical example cases were offered to show how the sheets were supposed to be completed (for similar research strategies cf. Schunk cit. in Tester 1999, 155). The data sheet forms were translated by the local research teams and distributed at the studied units; the local researchers also instructed the service staff in data collection: how to complete the sheets, and what interactions to include.

A total of 300 data sheets were collected. Each sheet recorded one episode of deviance, hence the number of different individuals is lower than the number of sheets because for some clients several incidents of friction were recorded. The studied population included various degrees of care dependency, from clients of social services getting one call a week to bedridden clients receiving continuous care. The variety of care dependency within the client population is related to the care policies operating in each country. If home care service is quite comprehensive (as in the UK National Health Service) severely care dependent persons can be treated at home who would need to move to a nursing home if they were living in Austria or Italy. Therefore the heterogeneity of the studied population is in fact an effect of national/regional care policy. By collating data of such different degrees of care dependency we can highlight the implicit exclusionary practices that would go unnoticed if we had only regarded one national/regional system at a time (Table 1).

Table 1
Number of data sheets and number of individuals per country

Participant observation of the practice of admission and service delivery of these service units was conducted and complemented by further observations of the services, which were not part of the central study. The observation was conducted by the—multilingual—Austrian and the Belgian research teams; all observers were researchers, not service personnel, and external to the studied units. Detailed fieldnotes in German, Dutch, and English document these observations.

In Austria the observation was conducted over 1 week periods in a unit providing home nursing and social service in November 2003 and February 2004 and complemented with ethnographic fieldnotes of an observation period of 8 weeks in December 2000 and January/February 2001. Ethnographic field notes of a period of 8 weeks collected in the Austrian day centre in August 1999 complemented this information.

The Belgian observation covered assessment house calls of social workers with cleaning and home help service in Turnhout in February and March 2004, as well as requests during office hours in community service social welfare centre in Dessel in March 2004, and requests during office hours and telephone contacts with clients in Turnhout, March 2004.

In Italy house calls with a home care service in Ancona and a day centre were observed over a period of 1 week in April 2004. Requests during office hours of the social worker in a local service district of Ancona were observed in September 2004.

In Belfast house calls of various facilities of the National Health Service (Rapid Response, Integrated Evening Service, Social Services, and Intensive Home Care) were observed by the author in March 2004 over a period of 1 week.

Records about statistics of user throughput, maxima of service hours, etc. were collected by the local research teams as well as brochures, information material, standards, and official documents of the services to serve as contextual information defining the range and scope of the studied services.

The possibly limited variance of cases of friction in the sample of data sheets was expanded by expert interviews (4–7 experts per country) representing comparable services in other regions of the studied countries.3 The interviews covered admission and discharge procedures, difficult clients, strategies of services in dealing with difficult clients, scope to accommodate requests and demands of clients, complaints and perceived expectations of clients and how the staff manage these.

In regular coding sessions the multilingual Austrian research team analysed the incoming data, generalized the codes and generated theoretical propositions about conditions of service use (cf. Charmaz 2000; Strauss and Corbin 1994, 1998; Glaser 1992, 1994) which led to the development of a theoretical model (see following section).4

To avoid misunderstandings it is necessary to clarify that the present case study is not meant to give a representative picture of four national welfare or service systems. For one thing this is not possible because our study aimed at generating knowledge about the nature and conditions but not the prevalence or frequency of exclusionary processes within service provision (Yin 2003, 10). It is also important to mention that due to regional disparities in all participating countries, the described practice may differ from the practice in other parts of the country, consequently the welfare state typology is too simple to capture these regional variations (cf. Anttonen et al. 2003).

Theoretical model

By generalising the reasons that service personnel identify as causes for friction, discharge or denial of admission, we collected evidence of explicit and implicit rules and conditions for becoming a user of social services. The resulting theoretical model describes the role of social care services in a mixed economy of welfare. It documents the general practice of admission, service provision, and discharge and highlights what reasons can prevent an older adult from becoming a service user.

The international comparison allows us to verify the inclusiveness of a service provision system independent of users’ traits since certain individual demands prevent service provision only in a given system of care provision (or in a particular organisational practice) (cf. Davey and Patsios 1999; Shea et al. 2003). Hence we can evaluate service systems and practices and identify potentials for optimisation. Above all, the theoretical model can be read in Hirschman’s terminology of exit and voice, and it can reveal how and to what extent choice and voice is tolerated and supported by a given care system.

The theoretical model (see graph) pictures the observed processes and marks critical paths (leading into precarious situations for individuals). The purpose of the model is to capture the data as consecutive events and to add information about logical conditions for service provision (and consequently exclusionary processes in service provision) that were observed (Fig. 1).

Fig. 1
Theoretical model: admission-service-discharge-cycle for one service provider (including an identification of potential exit and voice paths)

All possible variations of the case study can be described by the model (which does not mean that all conditions will necessarily occur in every studied country). A care dependent older adult enters the process on the left as a referral/request and consequently undergoes an admission procedure where her/his case is assessed for eligibility. After an older adult or his/her family directly contacts the service, either a head nurse or a social worker usually makes a house call in order to assess her/his needs in terms of care and social services, checks for resources such as family members or neighbours and assesses possible risks for care workers or the care process in general. Prior to this assessment a manager or staff on duty may pre-check—mostly by phone—whether the potential client is eligible in terms of age, operating area and overall expectations.

On the other hand referrals from hospitals, general practitioners, and social workers undergo assessment as a pre-selected privileged group with a low risk of rejection.

This admission/assessment procedure decides whether a private individual has a need, which can be addressed by the welfare system. In case of acknowledgement, the person could be referred to another provider or another service if the need is not being covered by this particular provider.

If the assessment accepts one of the needs being at possible risk of social exclusion in old age this transforms the older person to a ‘vulnerable old person’. Otherwise the need is rejected as irrelevant in terms of welfare services and no client relation is established.

Provider conditions

Being classified as vulnerable old person is a necessary but not a sufficient condition to be offered care services and to become a client. The vulnerable old person will need to meet other conditions, such as accepting service hours, service charges, and their care is subject to rules about the division of labour and fluctuation of personnel. A potential client will also need to comply with the provider’s expectations of adequate housing conditions concerning e.g., electrical safety or a—negotiable—level of cleanliness.

Service provision is also sometimes rejected by the potential client, particularly by persons with dementia who may then undergo further protection measures of the welfare state, such as institutionalisation or legal incapacitation if they put themselves at risk.

Shortage of staff on the side of the provider will prevent the establishment of a care relation and renders the vulnerable old person to a referral to another provider or a ‘virtual client’ by being put on a waiting list. Sometimes this search for a provider is done by the social worker or care manager, which saves the potential client from being assessed over and over again. The carousel of clients requesting help and being referred to another provider can have a discouraging effect on the clients who might eventually give up trying to get a service.

Continuous risk assessment during service provision

A vulnerable old person meeting all the conditions for service provision will become a client with a certain service package. The package may be adjusted according to changed needs.

While the client receives care services she/he may behave in a way, which can pose a risk to receiving a good quality of care.

If the client does not comply with diets or treatment plans or acts aggressively, insists unreasonably on a particular care worker or sexually harasses personnel, we speak of individual level risks originating in the particular behaviour of the client.5

On the family level conflicts between family of the care dependent and the care workers or family carers neglecting, exploiting or abusing the client can cause service users to become labelled as ‘difficult’ or ‘deviant clients’. For instance, staff reported clients, whose family would not provide clean underwear, or who were suspected of withholding proper food as ‘difficult’. This means that a service has difficulties dealing with clients who actually cause no friction at all themselves, but who have a family mistreating them or a family carer quarrelling with the provider.

The societal level risks originate in the social policy system and directly affect the client and her/his service, such as conflicts about costs or definitions of relevance of demands (e.g., weekend or night time service, or needs for socialising). Although a “[n]eed (rather than want) implies the existence of an objective condition” (Clarke 2006, 427) and is typically referring to inter-subjective concepts of necessity, the individual care service user who happens to express a demand which lies outside this collective agreement is seen as a nuisance to the service personnel. His/her non-compliance with the societal definition of legitimate wants is attributed to him/her as a personal fault.

From the care workers’ point of view all these deviations from the routine process of care provision create stress, and consequently clients bearing one or more of these risks tend to be avoided (Leppänen 2005). “Coping with troublesome clients in home care” (ibid.) includes talking about them and labelling them as ‘difficult’ and ‘deviant’ (from the provider’s routine) and thus applying rules and moralising about their behaviour (Becker 1973, 179; for the particular case of labelling in care relations see Vabø 2006, 411f.). Labelling and avoiding are two strategies for coping with ‘difficult’ clients. Care workers also have methods to appease or ignore deviant acts. They may for instance respect that a client has different standards of hygiene and not force him/her to shower every other day. They may excuse her/his unpleasant mood as a symptom of sickness. Also, the provider organisations have scope to make exceptions for clients. A risk which is exceptionally treated as if the person were complying with the rules—i.e., as if she/he were a co-operative client—loses its power to contribute to exclusion of the service.

So, depending on the reaction of the care personnel, a client with a risk can be defined as a co-operative client or a deviant client.

Deviant clients, those that have been labelled by the provider as difficult (e.g., because of aggressiveness towards care workers and nurses) will often receive the least necessary attention. They may for instance start having accidents and become incontinent, but the care workers might not interfere and offer assistance in such a delicate matter since they may fear that they could get shouted at or thrown out of the house. Only if the situation of the client deteriorates significantly, will the provider then act—and possibly at that point involve another provider. Deviant clients are permanently tested concerning their ability to cope with a minimum of help. If they succeed they remain deviant clients served by the provider (may be with special personnel assigned to care for them), otherwise they are forced to exit.

Co-operative clients have a better chance of getting an optimum of help or even special treats.

If more than one provider operates in a given area, deviant clients will become “hot potatoes” if they do not succeed in coping with a minimum of help, and they will consequently be passed to another provider, where they start with the admission and assessment procedure again. When passed to another provider the deviant client must start the negotiation process about tolerance of her/his particular demands all over again. It may well be that the client starts with a clean slate with the other provider, but it may also happen that the status of being a “hot potato” adds another risk.

Exiting the system

All clients—no matter whether deviant or co-operative—may exit the system either voluntarily or be forced out by one of the rules (e.g., a service hours maximum). The alternatives we observed after an exit are care by another provider, residential care, the black market, or informal care by the family.6

Exit and voice

In our study we observed that an equivalent to the Hirschmanian consumer exiting is a user changing the provider (often observed in Belgium), opting for informal care of family or the black market (in Austria the so-called “Slovak Nurses” and in Italy the “badanti”) or rejecting help altogether (observed in all countries except Italy, and often observed with clients of the UK National Health Service). Exit requires a substituting alternative. In this context it is questionable whether informal care is comparable—and thus fairly substituting for—the professional service.

The voice of a customer can be represented by friction observed in our study, caused by grumpy clients, clients rebelling against diets and prescriptions as well as demanding to be excepted from rules. Voice need not only be used by the client him/herself but can also be expressed by the family on behalf of the client.

Conflicts often follow rejected demands, e.g., due to rationing of services or prioritising tasks and can also be regarded as voice. For instance a bathing assessment in the UK National Health Service would allocate assistance for a bath in a tub only to persons with a skin condition, and offer assistive technology devices for all other clients. This was perceived as a decline in service quality by some clients who voiced their anger towards the care staff. Further, the explicit demands of clients or their family members that the professionals perform tasks in a particular way and to deliver the service at a particular time are direct responses—and thus voice—about perceived unsatisfactory quality to the provider. The same applies to clients or their families who expect the professional service to co-operate with and supervise black market care.

Limits to the effectiveness of exit

The model reveals that the power of the individual user to direct the service quality by exit is limited by some structural conditions of the care market.

1. Substitutability—are alternatives of similar quality available?

There are limits to exit, e.g., when two alternative producers offer goods of very different standards, a decline in quality of the initially better product will be accepted by the customers because the alternative is not really able to act as substitute. This relates in our example to the clients with a very high (medical) care need—who have no alternative to the professional service—and therefore cannot afford to exit. Similarly, Kremer found in a study on the use of personal care budgets in the Netherlands7 where care users are given cash to purchase services on the market, that “[l]ess professionalised help such as home care is more likely to be bought on the market. (...) It seems that patients in need of specialised care trust regular care much more, and professional nurses are more difficult to find outside the regular scheme” (Kremer 2006, 390).

2. Slackness of companies not interested in maximising efficiency

The traditional economic models assume that organisations and companies always strive to maximise their profit, and their effectiveness. Thus they would always deploy all available resources effectively in the production process. However, Hirschman contends that all organisations are affected by an inherent slackness, analogous to entropy in physics. This slackness is due to the fact that human society has increased productivity and control of the environment (and diseases) to a degree, which provides a certain buffer of security and does not require the full deployment of resources at any time.

The Hirschmanian premise of inevitable slackness in organisations serves as a valid generalisation of our observation of care providers: although the organisations we studied were created to deliver services which would prevent marginalization, many sacrifice this objective in deference to internal routines and labour law standards, the costs of service delivery, or similar secondary objectives. Optimising welfare services increases or maximises the necessary foundation of the political system—legitimacy (Schedler and Proeller 2003, 9)—but since this effect lacks immediacy for the service providers, they have no incentive to give their best. In this sense, welfare states are prone to slackness just like companies or organisations, and do not respond to exit (or voice) of their citizens. Waiting lists of the providers studied in Belgium lead to exits of clients from the care system: older persons who only need help with cleaning must wait up to 1 year in order to get the service, because the organization cannot hire the necessary staff due to lack of public funds. The output is not maximised and thus the goal of maximised social inclusion is not achieved.

3. Colluded competition (quasi conspiring competitors not really offering alternatives)

In the realm of the economy exiting customers may be replaced by new customers who are equally dissatisfied with any alternative supplier. If the quality of the product or service declines and the same happens with the products of the alternative producers, customers are forced to “shop around” bad alternatives without actually receiving better quality by exit. In addition, the producers do not have any incentive to improve the quality of their goods and services because their competitors are just as bad.

Competitors quasi-conspiring against their customers by fake competition and product diversification, confuse customers, and deprive them of the possibility of improving the quality of the product by the exit response.

In the realm of care we are confronted with a regulated market where quality is determined by laws, quantity is restricted by subsidised budgets and maxima of service hours, and the price is regulated. Consequently we could describe care systems with a host of different providers who basically all offer the same (potentially unsatisfactory) quality standard as colluded competition, which only produces the illusion to the client that they have a choice while at the same time they actually have no meaningful choice. Particularly in Austria and Belgium non-profit provider organisations of different religious or political backgrounds ‘compete’ for clients by offering essentially the same legally regulated service quality at the same prices, which does not represent competition in the care market.

4. Loyalty and induced loyalty (i.e., high initiation costs or entrance barriers)

Loyalty may restrict or block exit, because it produces additional costs for exiting. On the other hand, a loyal customer will be inclined to increase his/her influence on the management. Thus loyalty inhibits exit and activates voice (Hirschman 1974, 67). Only loyal customers will make threats to the management to exit. Other customers will exit immediately since this does not cause them any costs.

In the care systems loyalty is often expressed by clients who have a personal attachment to a particular care worker. The frequently reported conflicts about clients refusing to accept alternative care workers can be identified as loyalty which consequently activates voice—and thus friction in care provision—because clients who have a personal attachment, who feel loyal, are not put off by the costs of voice and they do not choose exit. Loyalty contradicts “the basic principle of consumerism” where the sovereign consumer does “not form trusting relationships with a service provider” (Vabø 2006, 405).

Loyalty may not always be the result of spontaneous personal choice but also be induced by institutional limits. This may be produced by high entrance costs/fees.

Waiting lists are a significant barrier to entering the system. In Hirschman’s terminology they represent high initiation costs. As Hirschman predicts, clients in this position will not immediately express their dissatisfaction (voice). Empirical data confirm this assumption of economic theory as an expert interviewed in Belgium says, “The real problem is here that elderly are really afraid to talk about their problems and how they would prefer to receive the aid. Because they are really dependent on the aid of the carer and they often have to wait first before they receive the aid. So they are glad to receive the help and they will wait a very long time to tell about the things they are not pleased with because they are afraid to lose help again. So when they do eventually they will rather be impatient and find it difficult to negotiate”.

Limits to the effectiveness of voice

1. Powerless voice in slack monopolies and demand outstripping supply

Hirschman questions the simplified concept of exploitative monopolies striving to maximise their profits, and claims that monopolies rather take advantage of their privileged market position in terms of undisturbed slackness and not in terms of maximised profits.

Under the condition where the supply does not meet the demand, the providers in competition act like slack monopolies: they ignore exits since they have more clients than they can actually serve. Furthermore the providers in this case also ignore voice and they are even happy to get rid of the most active critics (voice-customers exiting). This relates to the observation of so called hot potatoes passed from provider to provider. The strategy of passing on the hot potato saves the provider from adapting its service to the demands and criticism of the client. It preserves the comfortable slack climate of the organisation and expels the deviant client at the expense of reaching the goal of social inclusion. Consequently competition of the service providers is not always beneficial to the clients and does not empower them but rather the opposite.

The local monopolies providing formal care in Italy, Austria and the UK only allow exits to the informal sector of family care or black market care with the relative cost of the latter being much higher, in the context of a comprehensive public service system at low or no cost, such as the UK National Health Service.

2. Large number of isolated customers not aware of their impact on management

Voice requires that the customers are not dispersed and isolated but know of each other and can judge their relative power. This condition is not fulfilled in the case of the care system of an isolated client population. Particular programmes to stimulate the voice of empowered older persons have yet to be implemented. They will no doubt refer to universal citizenship rights to access public care systems independent of income, education or social status (cf. Vabø 2006, 418; Anttonen et al. 2003; Kronauer 2002, 152).

3. Costs of voice

Voice requires effort, time and skill to gather information and argue one’s case. Thus voice—and complaints such as a letter to the newspaper—are an option for more educated and less care dependent clients. Formalised complaint procedures of the care providers can have an important impact on voice, because they channel complaints and can improve quality if management takes the complaints seriously.

Some providers have signalled understanding and willingness to accept voice. “We have a slogan here that we have to see a complaint as a gift. So when we receive a complaint we can improve the quality of our assistance” (quality assurance officer in Belgium). Some systems include a standardised complaint management procedure. The UK National Health Service, as a monopolistic provider, encourages voice by standardised complaint management and reported dissatisfied clients contacting their Member of Parliament as well as a client being able to sue the National Health Service while receiving its service.


Applying Hischman’s economic theory about the limitations of exit and voice in the free market on the concrete empirical example of the care market suggests that we should be sceptical of the contention that a mixed welfare economy is able to increase autonomy and enhance the rights of service users in long-term care. The specific practice of care provision—as was observed in our study of deviant clients—has characteristics that prevent the empowerment of the users by choice.

Firstly, the imbalance of power between client and provider leads to the provider regulating the access to the market as can be seen in the theoretical model where assessment procedures and explicit and implicit rules and conditions restrict the user. The power a user can exert with her/his voice given that she/he can threaten with exit in a mixed economy of welfare is minor compared to the provider organisations’ power to label criticism as recalcitrance and to expel a critical customer as a “hot potato”. The latter can happen even more easily when competitors of the provider are available. Providers do not just execute rules and regulations set by the framework of social policy but have been observed to make exceptions, which means that the rules are negotiable.

Furthermore, care provision is often operated by local monopolies or under the imbalanced market condition of under-supply, the latter creating ‘hot potatoes’ of ‘difficult clients’ passed from one competing provider to the other. Similarly Kremer reports about the Dutch system that some provider organisations reject personal care budget holders as clients because their individualised demands cause too much friction to the organisation of care service provision (Kremer 2006, 391).

Secondly, the care market is regulated to the extent that professional services must comply with health and social care legislation and are offered at fixed prices. It is questionable to what extent colluding competitors—such as Christian charity or a socialist non-profit provider—are offering an alternative to a dissatisfied client. Although the concept of the mixed economy of welfare includes informal care, exit from the professional service system to the informal care of families or the black market can hardly be counted as exit to an equivalent alternative. The two alternatives are different both in quality and quantity of the service.

Finally, ‘distributive effects’—as economists call biased demand—distort the individual freedom where the care market offers choice and where the clients also use this choice. Choice in general is used in the interest of price consciousness at the expense of quality.

This means that within a full range of products of different quality and prices, all consumers except those at the upper and lower margin (who do not have an alternative of full substitutability) will start to exit to the closest acceptable alternative when the quality of their product declines. However, consumers with high quality standards who can afford higher prices will tend to choose better and more expensive alternatives and thus get higher benefits. Consumers with more consciousness for prices (consumers who cannot afford the upgrade) will tend to the cheaper alternatives of lower quality.

Similarly to the recipients of the cash benefit of the long term care insurance in Germany, a group in which the lower classes are overrepresented (Blinkert and Klie 1999 cit. in Theobald 2004, 23), care dependent persons receiving cash care allowance in Austria and Italy often prefer not to spend this money on qualified care—but simply add it to the household budget (Ungerson 2003, 2004). The price conscious—and poorer—recipients of the cash benefits do not use the care allowance to hire qualified personnel delivering a service of a given and claimable quality standard but settle with cheaper solutions of a neighbour or a ‘badante8 (i.e., a black market carer in Italy). Our data about the ‘deviant clients’ document frequent conflicts with clients about costs, which evidences the price consciousness of care dependent older adults and their families.

We have observed yet another example for these distributive effects in our study: many wealthier family carers make an effort to get the best available medical products and aids for the care. Poorer clients and family carers often shun the effort and costs of, e.g., a proper care bed or pressure mattresses and accept the risk of back injuries to the carer or sores to the care dependent instead.

A universalistic care system—based on public responsibility—seems to encourage and acknowledge voice more effectively, as our evidence about the National Health Service in Northern Ireland shows. The users can claim citizenship rights on the service and exert legal and political pressure by involving their Member of Parliament or by writing letters to the newspapers. The services usually have established formalised complaint management processes and thus facilitate voice as opposed to exit. The social status and education of the dissatisfied client influence the impact of their voice. Older adults from lower social classes tend to be disadvantaged. The introduction of market mechanisms is not only unable to balance this disadvantage, but rather increases it.

Although the present case study can not claim to represent a complete picture of the role of choice in welfare economies, its insights and the specifically economic interpretation of the data in Hirschmanian terms may be viewed as an empirically grounded contribution to the ideological discussion of the advantages of market mechanisms in the delivery of welfare services.


The following paper is an outcome of the research project CARMA (Care for the Aged at Risk of Marginalization) which was funded by the European Commission within the Fifth Framework Programme (QLK6-CT-2002-02341). The author is solely responsible for the content, which does not represent the opinion of the European Community; the Community is not responsible for any use that might be made of data appearing in this paper. I would like to express my gratitude for the excellent collaboration with my research partners Roswitha Baumgartner (A), Francesca Cesaroni (I), Guido Cuyvers (B), Nadia Gentile (A), William McCormick (UK), and An Pintelon (B). I am also indebted to Norman Alm (University of Dundee), Christian Fleck (University of Graz), Stefan Laube (University of Konstanz), William McCormick (NHS Belfast) and the two anonymous reviewers for their valuable comments that helped me revise and improve the paper.


An external file that holds a picture, illustration, etc.
Object name is 10433_2007_47_Figa_HTML.jpg


1The research project CARMA (Care for the Aged at Risk of Marginalization)—of which the present study was a part of—studied what service offers are made to older care dependent persons by the welfare state and to what extent these services prevent marginalization. Albeit a number of publications of the project have dealt with the concept of social exclusion in old age, it would go beyond the scope of this paper to discuss it here (cf. Theobald 2005a, 2005b, 2006a, 2006b, 2007; Rauhala 2003; Egger de Campo et al. 2005b). The multidimensionality and the cumulative nature of social exclusion in old age are core concerns for the CARMA study (cf. Kronauer 2002, 17f.). In addition, the exclusion from social rights—such as access to welfare services addressing persons in need—represents the particular focus of the study reported in the present paper (Kronauer 2002, 185).

2The studied home nursing or home care services for older people were located in Graz (Austria) in Turnhout, Dessel, Balen, and Antwerp (Belgium); in Ancona (Italy) and in Belfast (Northern Ireland). The two day centres for seniors were in Graz and Ancona. Apart from the Northern Irish services all were private non-profit organisations commissioned by the public authorities.

3These interviews were conducted by the local research teams between August and October 2004 in Austria, in September 2004 in Belgium, between October and December 2004 in Italy, and in July 2004 in Northern Ireland.

4A detailed—200 pages—description of the codes and steps of analysis can be found in Egger de Campo et al. 2005a.

5For a full report of the various risk categories please refer to Egger de Campo et al. 2005a.

6The cases where the clients die or recover or move out of the operating area have not been considered because they do not describe a social process relevant to social exclusion or choice in a market economy

7 Persoonsgebonden Budget

8It must be said, however, that the badanti very often complement the insufficient supply of the professional care system.


  • Anttonen A, Sipilä J, Baldock J. Patterns of social care in five industrial societies: explaining diversity. In: Anttonen A, Sipilä J, Baldock J, editors. The young, the old and the state social care systems in five industrial nations. Cheltenham: Edward Elgar Publishing; 2003. pp. 167–197.
  • Baldock J. On being a welfare consumer in a consumer society. Soc Policy. 2003;2:65–71. doi: 10.1017/S1474746403001088. [Cross Ref]
  • Becker HS. Labelling theory reconsidered. In: Becker HS, editor. Outsiders. Studies in the sociology of deviance. New York: The Free Press; 1973. pp. 177–212.
  • Blackman T, Brodhurst S, Convery J. Social care and social exclusion. A comparative study of older people’s care in Europe. Palgrave/Houndmills/Basingstoke/Hampshire: Macmillan; 2001.
  • CARMA (2005) Care services for the elderly: state of the art and perspectives. Proceedings of the seminar Fano, 4th November 2004. CooSS Marche, Ancona
  • Charmaz K. Grounded theory. Objectivist and constructivist methods. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. 2nd edn. Thousand Oaks: Sage; 2000. pp. 509–535.
  • Clarke J. Consumers, clients or citizens? Politics, policy and practice in the reform of social care. Eur Soc. 2006;8:423–442. doi: 10.1080/14616690600821966. [Cross Ref]
  • Davey A, Patsios D. Formal and informal community care to older adults: comparative analysis of the United States and Great Britain. J Fam Econ Issues. 1999;20:271–299. doi: 10.1023/A:1022957426159. [Cross Ref]
  • Denzin NK. The research act: a theoretical introduction to sociological methods. New York: McGraw-Hills; 1970.
  • Egger de Campo M, Baumgartner R, Just B (2005a) Evaluation of conference instructive deviance – interaction between users and service providers. Deliverable 13 of CARMA—Care for the Aged at Risk of Marginalization. QLK6-CT-2002-02341, Graz, download
  • Egger de Campo M, Baumgartner R, Just B (2005b) Lehrreiche Abweichungen. Ergebnisse der CARMA Fallstudie (Instructive deviance. Results of the CARMA case study). In: Egger de Campo M, Posch K (eds) Strategien gegen soziale Ausgrenzung alter Menschen, Erfahrungen aus der Praxis (Strategies against social exclusion of older people. Lessons from practice). Graz, pp 19–60
  • Evers A. The mix of care provisions for the frail elderly in the federal republic of Germany. Deficits, changes and prospects of reform. In: Evers A, editor. New welfare mixes in care for the elderly. Vienna: European Centre for Social Welfare Policy and Research; 1991.
  • Finer CJ. Trends and developments in welfare states. In: Clasen J, editor. Comparative social policy. Concepts, theories and methods. Oxford: Blackwell; 1999. pp. 15–33.
  • Folbre N. The invisible heart. Economics and family values. New York: The New Press; 2001.
  • Glaser B. Emergence vs. forcing. Basics of grounded theory analysis. Mill Valley: Sociology Press; 1992.
  • Glaser B. More grounded theory methodology. A reader. Mill Valley: Sociology Press; 1994.
  • Hirschman AO. Exit, voice, and loyalty: responses to decline in firms, organizations, and states. Cambridge: Harvard University Press; 1970.
  • Hirschman AO. Abwanderung und Widerspruch (Exit and voice) Tübingen: J.C.B. Mohr; 1974.
  • Kremer M. Consumers in charge of care: the Dutch personal budget and its impact on the market, professionals and the family. Eur Soc. 2006;8:385–401. doi: 10.1080/14616690600822006. [Cross Ref]
  • Kronauer M. ‘Soziale Ausgrenzung’ und ‘Underclass’: Über neue Formen der gesellschaftlichen Spaltung (Social exclusion and underclass: about new forms of social division) Leviathan. 1997;25:28–49.
  • Kronauer M. Exklusion. Die Gefährdung des Sozialen im hoch entwickelten Kapitalismus (Exclusion. Jeopardizing the social in highly developed capitalism) Frankfurt/New York: Campus; 2002.
  • Leppänen V (2005) Coping with troublesome clients in home care. Paper presented at the ESA (European Sociological Association) meeting in Torun, Poland
  • Lundsgaard J. Choice and long-term care in OECD countries: care outcomes, employment and fiscal sustainability. Eur Soc. 2006;8:361–383. doi: 10.1080/14616690600821974. [Cross Ref]
  • Mabbett D, Bolderson H. Theories and methods in comparative social policy. In: Clasen J, editor. Comparative social policy. Concepts, theories and methods. Oxford: Blackwell; 1999. pp. 34–56.
  • Rauhala M (2003) Deliverable 1 of the project CARMA: comprehensive description of the social process of marginalization of the aged according to social scientific literature. (Obtainable as download from:
  • Rake K. Accumulated disadvantage? Welfare state provision and the incomes of older women and men in Britain, France and Germany. In: Clasen J, editor. Comparative social policy. Concepts, theories and methods. Oxford: Blackwell; 1999. pp. 220–245.
  • Rostgaard T. Constructing the care consumer: free choice of home care for the elderly in Denmark. Eur Soc. 2006;8:443–463. doi: 10.1080/14616690600822048. [Cross Ref]
  • Schedler K, Proeller I. New public management. Bern/Stuttgart/Wien: UTB; 2003.
  • Shea D, Davey A, Femia AA, Zarit S, Sundström G, Berg S, Smyer MA. Exploring assistance in Sweden and the United States. Gerontologist. 2003;43:712–721. [PubMed]
  • Sipilä J, Anttonen A, Baldock J. The importance of social care. In: Anttonen A, Sipilä J, Baldock J, editors. The young, the old and the state. Social care systems in five industrial nations. Cheltenham: Edward Elgar Publishing; 2003. pp. 1–23.
  • Strauss A, Corbin J. Grounded theory methodology: an overview. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. Thousand Oaks: Sage; 1994. pp. 273–285.
  • Strauss A, Corbin J. Basics of qualitative research. techniques and procedures for developing grounded theory. Thousand Oaks: Sage; 1998.
  • Tester S. Community care for older people. A comparative perspective. Houndsmills/Basingstoke: Macmillan; 1996.
  • Tester S. Comparative approaches to long-term care for adults. In: Clasen J, editor. Comparative social policy. Concepts, theories and methods. Oxford: Blackwell; 1999. pp. 136–158.
  • Theobald H (2004) Care services for the elderly in Germany. Infrastructure, access and utilisation from the perspective of different user groups. Berlin. Veröffentlichungsreihe der Arbeitsgruppe Public Health Forschungsschwerpunkt Arbeit, Sozialstruktur und Sozialstaat. Wissenschaftszentrum Berlin für Sozialforschung (WZB)
  • Theobald H (2005a) Social exclusion and social care for the elderly. Theoretical concepts and changing realities in European welfare states. Discussion-paper (SPI 2005-301), Social Science Research Center (WZB), Research Group Public Health, Berlin
  • Theobald H (2005b) Social exclusion in old age: challenges for elderly care approaches. In: Divay S (ed) Welfare and caring professions. Special issue knowledge, work, society, vol 3. pp 41–57
  • Theobald H (2006a) Restrukturierung der Altenbetreuung. (Restructuring Elder Care) In: Rehberg KS (ed) Soziale Ungleichheit, Kulturelle Unterschiede. Verhandlungen des 32. Kongresses der Deutschen Gesellschaft für Soziologie in München 2004. (Proceedings of the German Sociological Association) CD-Rom. Frankfurt/Main Campus, pp 3700–3709
  • Theobald H (2006b) Care resources and social exclusion: a European comparison. In: Backes G, Lasch V, Reimann K (eds). Gender, health and ageing. European perspectives on life course, health issues and social challenges. VS Verlag für Sozialwissenschaften, Wiesbaden
  • Theobald H (2007) Social exclusion, social integration and care: concepts and empirical findings in European countries. In: Künemund H, Schroeter K (eds) Social inequality and cultural differences during the life course and in old age—facts, prognosis and visions. Series Age Soc 12, VS-Verlag, Wiesbaden (in press)
  • Ungerson C. Social politics and the commodification of care. Soc Polit. 1997;4:362–381.
  • Ungerson C. Commodified care work in European labour markets. Eur Soc. 2003;5:377–396. doi: 10.1080/1461669032000127651. [Cross Ref]
  • Ungerson C. Whose empowerment and independence? A cross-national perspective on ‘cash for care’ schemes. Age Soc. 2004;24:189–212. doi: 10.1017/S0144686X03001508. [Cross Ref]
  • Yin RK (2003) Case study research. Design and methods, 3rd edn. Sage, Thousand Oaks
  • Vabø M. Caring for people or caring for proxy consumers? Eur Soc. 2006;8:403–422. doi: 10.1080/14616690600821990. [Cross Ref]
  • Walker A. The social creation of poverty and dependency in old age. Soc Policy. 1980;9:49–75.
  • Walker A. Towards a political economy of old age. Age Soc. 1981;1:73–94. doi: 10.1017/S0144686X81000056. [Cross Ref]

Articles from European Journal of Ageing are provided here courtesy of Springer