1In the early twentieth century, the development of a public sphere employing a growing share of wage-earners led a number of sociologists to reflect on the values and practices current in this world, as opposed to those in the private sector and among the self-employed. Of the discipline’s founding fathers, Max Weber ( 1971, p. 222) contributed greatly to the thinking about this distinction, highlighting the specificity of actors who act in the name of bureaucratic rationality. Robert K. Merton (1952), meanwhile, attempted to establish a link between bureaucratic structures and the personalities of those who compose them. But it is especially in the debates on the transformation of social classes that the divide between public and private has given rise to controversy.
2At the end of the 1970s and throughout the 1980s many sociologists pointed out the pertinence of this division for contemplating the differences between class factions. In France, Pierre Bourdieu had already drawn attention to differences in tastes between managers in the public and private sectors (Bourdieu, 1979, p. 130); and a few years later, François de Singly and Claude Thélot’s statistical description brought to light the various forms this “great difference” could take, either in social and occupational trajectories or in lifestyles (Singly and Thélot, 1988). Comparable perspectives have been developed in Scandinavian countries where various authors have advocated analysing public employees as a specific group (Hoff, 1985; Hoff and Andersen, 1989). Rejecting the term class in its Marxist sense, Jens Hoff considers that the public sector has its own form of organization, and that this is relatively independent of the private sector. Others go further and support the thesis that in Norway, Sweden, and Denmark the public sector middle class forms a specific social class, which they name the “new class” (Hoel and Knutsen, 1989). Conversely, in the United Kingdom, work by John Goldthorpe led to a prevailing view of the social world that took no account of this difference: in his classification, managers and professionals are amalgamated in the “service class” category, encompassing business executives, administrative managers and technical specialists. The argument often advanced for not treating public and private sector workers differently is that the differences between the working, middle and upper classes are much more significant than those relating to status or employment sector. Nevertheless, there is a significant strand in British sociology for which the opposition between public and private is a pertinent descriptive criterion for the transformation of the middle class.  These sociologists study the 1980s and the Thatcher years as a turning point during which the public–private divide became more important than the social class divide (Edgell and Duke, 1991). With the arrival of the Conservatives, the public sector was threatened by the growth of the private sector and it also became a central issue in the conflicts throughout British society. For example, public sector managers had different interests to those in the private sector who took over functions the former had hitherto carried out.
3The public–private divide rarely appears as a criterion for describing social groups in studies comparing the social structures of different European countries. To show that class variations still have explanatory power, most sociologists and statisticians adopt the hypothesis of the so-called EGP model (in reference to Erikson, Goldthorpe and Portocarrero, 1979), which favours an aggregate grouping level based on employment relations (Erikson and Goldthorpe, 1992). Two principal arguments are advanced to justify not taking (public or private sector) status into consideration in the definition of class position. Firstly, the differences between states concerning what they consider to be the “public sector” are too significant (Rose, Pevalin and Elias, 2001, p. 81). Belonging to the public sector, moreover, does not have the same meaning in every country. France remains emblematic of a career civil service that is characterized by the central importance of competitive examinations and a guarantee of lifelong employment. In Scandinavian countries and Great Britain, public services since the 1990s have been organized along the lines of a jobs-based civil service, in which civil servants are recruited through job interviews, with no competitive examinations (Eymeri-Douzans, 2012, p. 324). However, despite the different organizational characteristics, European comparisons indicate that belonging to the public sector fosters greater job security (Dieckhoff, 2011) and involves a different type of relationship to work to that of private sector employees. Secondly, those who reject the relevance of employment sector cite the observation that the New Public Management turn accelerated the increasing similarity of employment and work conditions between the public and the private sectors. While this process is now well documented in France (Torterat, 2009; Guillemot and Jeannot, 2013), it does not necessarily imply that public sector workers share the same views and adopt the same practices as those in the private sector.
4To assess this, we have chosen to focus on work addressing the public–private divide at the European and national levels during the years 1990–2000, focusing our attention on four countries: Germany, France, Great Britain, and Sweden. There are three principal reasons for this choice. Firstly, these countries carry significant weight within the European Union because of the size of their work forces and the state models they embody. Next, while they have been shaped by different histories in terms of state construction, they have public sectors that play large roles and employ a significant portion of their wage earners. Finally, in these four countries, debate on social inequality has remained very dynamic, in particular for assessing different types of social mobility in Europe (Erikson and Goldthorpe, 1992). The cost of this generalization is, however, an obscuring of the division between public and private sector workers that remains highly significant in many works studying social structures at the national scale. In Great Britain and Sweden the debate focuses on differentiations within the middle class, while in France and Germany it is more general. This contradiction between comparative studies that obscure the role played by employment sector and national studies that attribute great importance to it invites reflection on the social characteristics shared by public sector employees: relationships with the state, functions and futures that are distinct. It is undoubtedly in the realm of political opinions and trade union practices that this divide is most pertinent, but it can also lead to differences in relationships to work and cultural practices and also in consumer practices.
5This literature review is also an opportunity to revisit the debates that, since the 1990s, have focused on the existence of a common work ethos among public sector employees. We propose linking this to the notion of habitus, which, beyond professional values, involves thinking of the representations shared by public sector employees as “principles which generate and organize practices” (Bourdieu, 1980, p. 88) in cultural, political, and economic fields. This move seeks to emphasize the specificity of the “public sector middle class,” and, in particular, the link between representations and dispositions incorporated by these employees and their position in the social hierarchy. The managerial reforms of states in Europe have redrawn the public sector boundaries, confirming a sharper divide between managers and middle managers or subordinate agents. The aim is not to substitute a distinction based on employment sector for an analysis in terms of social class. Our intention is instead to show that in many European social science studies, the opposition between public and private sectors remains a structuring, albeit secondary, dimension of social hierarchies.
The European public sector halo
6In Europe, the boundary that separates the public and private sectors is often blurred and varies considerably depending on a state’s development and the pace of administrative reforms. It is more evident in countries where the civil service retains its influence and where its status is constitutive of very pronounced differences (as in France and Germany). It is conversely much more porous in countries such as Great Britain and Sweden where liberal reforms have considerably reshaped the functioning and scope of state intervention through privatizations and by the expansion of executive agencies. This divide overlaps in part with that which opposes a career civil service, composed of closed jobs structured by professional bodies, with a jobs-based civil service made up of posts made available on the basis of career background (Bezes, 2008, p. 224).
7Depending on which criterion one adopts—status of employment, the public or private nature of the employer, or how employees identify themselves—the contours of the population one means to study vary considerably. To reduce these taxonomic ambiguities, we have chosen to distinguish statutory civil servants frp, agents whose employment fulfils “public interest functions” (Audier and Bacache-Beauvalet, 2007). This approach enables us to compare public employment in all European countries, retaining the idea that it is a halo made up of a hard core of statutory posts with peripheral strata offering very varied contracts from one country to another.
Statutory civil servants
8In the sense used by Community Law, the status of civil servant is defined by two criteria: the exercise of powers conferred by public law and the responsibility to defend the public interest of the state or civil society (Demke, 2004, p. 100). In most European countries this means, a minima, jobs corresponding to certain positions or missions relating to national sovereignty: judges, military personnel, tax officials, customs officers, police, or ministerial employees. But beyond that, the boundary between statutory civil servants and public officials varies from one country to another: in some countries the former can represent up to 90% of public employees and in others barely 15%. The administrative structure also varies: French and German ministries have large numbers of agents to formulate and apply policies countrywide. In contrast, in the United Kingdom and Sweden, ministries are mostly smaller and are devoted to the formulation of public policies that are carried out by agencies.
9In France, most public agents—83% in 2009 (DGAFP, 2011)—have civil servant (fonctionnaire) status, an employment status that is very different from the private sector in terms of employment security, remuneration, promotion and retirement (Friot, 2012). They are mainly split between the civil service of the state (fonction publique de l’État), the civil service of local governments (fonction publique territoriale), and the civil service of public hospitals (fonction publique hospitalière); each of which contains “corps” and grades that correspond to determined levels of remuneration.
10The civil service in Germany is older than most other European countries: its status (Berufsbeamtentum) emerged in the eighteenth century and is rooted in Prussian history. According to the traditional view of the Obrigkeitstaat, German civil servants are in the service of the state, which involves certain protections and obligations (Derlien, 2003): they are bound by a duty of neutrality and loyalty to the Constitution, they cannot strike but are guaranteed a job for life and are expected to make their career in the service of the public authority. In 2003 there were 1.7 million statutory civil servants, which represented about 40% of employees in the public sector (Bossaert and Demke, 2005, p. 32).
11At first sight, Sweden appears to have the largest public sector (employing 31% of the active population). But this observation is tempered by the fact that the vast majority of these agents are governed by rules and legal dispositions that do not fundamentally differ from their counterparts in the private sector. Less than 1% of public sector workers have civil servant status, most of whom are judges (Demke, 2005, p. 33). They are not recruited through competitive examinations and since the 1960s public law rules have contributed to a convergence between public service and the private sector, particularly in terms of remuneration linked to performance and merit (Modern, 2004).
12In Britain, the civil service has a very different meaning: far from being purely a state intervention instrument, it is seen as mediating between the interests of different social groups (Knell, 2001, p. 73). Despite early industrialization, the development of a welfare state occurred later here than it did in Germany (Esping-Andersen, 1990). Since there is no common civil servant status, the means of recruitment and promotion correspond to an employment system and not a career. The administration is organized at two levels: central government, which includes the ministries and armed forces, and local government, which encompasses education, social services and the police. Considered “servants of the crown,” civil servants can be removed from office by political decisions, but their careers are relatively protected by the legislature that grants them specific rights. They represent only 10% of public sector employees, approximately 550,000 agents (Bossaert and Demke, 2005, p. 32). Excluded from this number are many agents carrying out public interest functions such as judges and police officers. Many health and education professionals are employed by executive agencies.
13The persistence of a core of statutory civil servants has not prevented the rules they must abide by from evolving with the spread of New Public Management: the introduction of targets, variation of remuneration depending on performance, encouragement of continuous education, and mobility within organizations (Emery and Giauque, 2005, p. 686).
14More and more public service functions in Europe are fulfilled by agents who do not enjoy guaranteed lifelong status, but work in more or less stable employment situations. To distinguish them from statutory civil servants, we have chosen to call these people public agents.
15In France, the public agents category refers to those who do not belong to the state civil service, which in 2010 represented 22.4% of the work force in the public sector (principally in the local government civil service and the public hospitals civil service), but in the last ten years their number has risen twice as fast as that of statutory civil servants (2.8% against 1.3% per year). These public agents have various statuses, defined by public and private law, from mission heads to temporary agents, through all forms of auxiliary staff. Some may eventually be granted permanent contracts through policies to reduce precarious employment, the extent of which varies depending on the period in question.
16In Germany, public agents employed on a contractual basis are more often employees and manual workers who usually work in the social and health sectors and for local government, that is for communes or groups of communes (Gemeinden/Gemeindeverbände). They do not automatically spend their whole career in the same jobs, are subject to the collective conventions of the private sector and can be dismissed. Their salaries can be negotiated, and depend not on their grade but on the job they are doing. Moreover, while civil servants’ superannuation is the state’s direct responsibility, there is a federal insurance fund for employees (Bundesversicherungsanstalt für Angestellte) and different Länder have different insurance agencies (Landsversicherungsanstallten) for workers. Nevertheless, some public agents can perform the same job for a long time with a salary progression comparable to statutory civil servants. How can we tell if a newly recruited person should be a civil servant or a salaried employee under private law? There seems to be no clear rule: within the same institutions, teachers can either be employed as Beamte or as Angestellte (Keller, 1999, p. 57).  There has also been a rise in part-time employment: this now represents a quarter of the German public sector work-force, most of whom are employees and workers working in the Länder.
17In Sweden, the public agents’ employment terms tend to come close to, or even to merge with, those in the private sector. The historic 1965 agreement making it possible for trade unions to negotiate collective agreements constituted the beginning of the replacement of the single wage scale by a remuneration system that is similar to the private sector (OECD, 1993, pp. 121–2). Indeed, while the principle of general salary scales persisted until the end of the 1990s, Swedish public authorities have gradually gained the discretionary ability to adjust the relationship between a position and a grade depending on the competence and aptitude of employees. The 1986 decentralization law then led to the disappearance of the collective salary scale and resulted in 1989 in a framework agreement which organized Swedish public sector employee remuneration on the private sector model (ibid., pp. 123–4). The signing of an “agreement on job security” thus now enables public authorities to dismiss some civil servants (with the exception of judges and the armed forces), provided that they give them notice and offer them redeployment and training assistance. For the other public sector employees, the power to recruit and define working conditions essentially rests with the municipalities (Kommuner) and regional institutions (Landsting) whose powers were extended by several reforms between the end of the 1980s and 1990s. In 2004, it was estimated that 83% of Swedish public sector employees were in local government (Biland, 2012, p. 21).
18In Britain, public sector agents include all people employed by local authorities (the local government service) and by executive agencies under the supervision of central government. They are recruited on private law contracts and subject to general labour legislation in the same way as private sector employees. Until recently, some were recruited to public corporations—the equivalent of French public companies—but these for the most part have disappeared with the privatizations in the 1980s (telecommunications, management of ports, aerospace, railways, etc.). New organizational models that are similar to the private sector have been developed, for example the PPP (Public–Private Partnership) and executive agencies. Finally, a vast decentralization programme has led to public sector employment more often being managed by local authorities rather than by central government (Bach and Winchester, 1999).
Withdrawal or redeployment of the public sector?
19The public sectors in all European countries have been affected quantitatively and qualitatively by the New Public Management reforms. The trend towards a reduction of the number of civil servants can be seen as a massive decline of the public sphere in Europe or as a redeployment of statuses associated with public service missions.
20In Germany, the public sector was completely reorganized following the country’s unification which led to a substantial shrinkage particularly among the Länder in the East: the number of agents working for public authorities declined from 6.4 million in 1991 to 4.2 million in 2003.
21In Britain, the lack of a unified public service status makes it difficult to measure the extent of public employment.  Nevertheless, it can be estimated that, between the 1980s and 1990s, public sector employment fell from 7 million employees (30% of all employees) to a little more than 5 million (22%) (Roberts, 2001). This decline was particularly sizeable in the public corporations that were privatized (500,000 jobs passed from the public to the private sector). While the share of public sector employment within total employment stabilized at around 20% during the 2000s, the number of public employees has risen from a little more than 5 million in the mid-1990s to 5.8 million in 2005, notably because of the significant rise in numbers in the National Health Service (Hicks et al., 2005). The British public service reform process thus led to a reduction in the scope of the public sector, but this process masks growth in the number of public service related jobs: in the 1980s and 1990s the share of employment relating to public interest roles rose from 24% to more than 30% (Gaudier and Bacache-Beauvalet, 2007). These are mostly education, social and health roles that were redeployed during this period, upsetting the differentiations between middle-ranking and senior employees (Roberts, 2001; Crompton, 2008).
Box 1.—The public–private divide in national classifications
22Although the European public sector has undoubtedly been transformed over the past three decades, it would be wrong to concern ourselves only with the reduction in its size. The decline in bureaucracy has also occurred through the receding of a certain type of administrative authority in favour of greater weight being given to skills and thus cultural capital (Savage et al., 1992; Butler and Savage, 1995). In this perspective, the reduction in the public sector is accompanied by a rise in the power of professionals, who have seen their incomes increase significantly more than other occupational groups. The decline in the number of public sector workers has been compensated for by the development of the social, education and health sectors. And while the state is not always in the position of direct employer, it continues to intervene as a purchaser of services. This is particularly the case for professionals whose autonomy and power is founded on state intervention (certification and diplomas, regulation of these sectors, welfare spending, etc.) (Savage et al., 1992, pp. 72–6).
Differences that persist in the occupational sphere
23We now want to address the significant differences that persist between the public and private worlds, drawing on sociological studies devoted to this question and based on quantitative statistics. Many researchers have described a convergence between public and private practices in the area of working conditions, even suggesting a continuous process of hybridization between the two spheres (Emery and Giauque, 2005, p. 690). However, the growing convergence between the two sectors from the point of view of working conditions does not contradict the persistence of significant differences that are evident as much at the European level as at the national scale.
More stable and less unequal employment conditions
24Although governed by different legal regimes, the public salaried labour force in European countries was formed outside the rules of the labour market, notably on the principle of greater job security—considered firstly as a protection against politicians at the head of governments—and a remuneration system dependent on agents’ statuses and qualifications and not on the wealth they generate or the work they accomplish (Weber,  1971, p. 22).
25The divide linked to job security continues to be relatively pertinent today. According to data from the European Community Household Panel, public employees are, unsurprisingly, less likely than others to experience unemployment (Dieckhoff, 2011, pp. 240–1). To this must be added the different resources that these jobs provide. In the United Kingdom, access to continuing education is much greater in the public than in the private sector; public sector employees also benefit from more health cover and better pensions (Bach and Winchester, 1999; Meyer and Bridgen, 2008). The greater likelihood in the public sector of having access to higher education or pensions benefits is confirmed by other studies and for a wider range of European countries (Oesch, 2006, pp. 155, 206).
26In terms of employment conditions, studies comparing the public and private sectors have mostly focused on differences in incomes and salaries, what economists call the public sector pay gap. There are differences between states in this area. In this sense, France is characterized by a progressive disparity between public and private sector employees due to a slowdown in pay rises at the bottom of the hierarchy and a relative depreciation of remuneration among the highest categories (Baudelot et al., 2014, p. 49). According to a study by the European Central Bank, this is not an isolated case. Between 1995 and 2009 the ratio between public and private sector salaries was slightly less than 1 in France, Germany and Belgium, while it was greater in Austria, Spain, Italy, Greece, Portugal, Ireland, and Slovenia. Moreover, measuring it is complex since effects linked to socio-occupational structures and qualifications in the two sectors have to be controlled for (Giordano et al., 2011). All things being equal, moreover, some categories of public employees have a salary advantage over those in the private sector, namely women and agents at the bottom of the salary scale (Lucifora and Meurs, 2006; Giordano et al., 2011). However, this advantage tends to diminish when monthly salary is taken into account—rather than hourly wage—and if we restrict the comparison to employees in large private firms. Furthermore, these gaps are very comparable regardless of whether the countries are well advanced in terms of state reform or not (Giordano et al., 2011). In other words, the managerial turn does not seem to have brought the public and private sectors closer together in terms of remuneration conditions. From these studies on wage differentials, we should note a structuring element in the differing representations between these two sectors: remuneration levels are less unequal in the public than in the private sector. Thus, there is a narrower salary range (Oesch, 2006, p. 113) and the gaps between men and women are less pronounced (Meurs and Ponthieux, 2005). 
The social characteristics of public sector workers
27Differences between the public and privates sectors are also observable when we compare the social morphology of employees. The first structural characteristic of the public sector concerns the proportion of women. Women have greatly benefitted from the expansion of public sectors (Kolberg, 1991), but employment rates vary greatly from one country to another: in 2010, they represented 52% of the public sector in Germany and 65% in the United Kingdom (according to ILO figures).  Beyond this significant presence, there are still fewer women than men in management positions.
28In France, the feminization of employment in the public sector is significant (59%), but nevertheless varies depending on government authority (state civil service, public hospitals, and local government) and social category. When we observe the gender distribution of different categories of employees in the public service, we thus find an area that is majority feminine (more than 60% women) encompassing health and social occupations (“Intermediate occupations” [“Professions intermédiaires”]), those in primary education (“School teachers and auxiliaries” [“Professeur des écoles et assimilés”] and service agents (“Employees” [“Employés”]) and an area occupied almost completely by men, like police and military jobs and manual workers (fewer than 20% women) (Hugrée, 2011, pp. 70–1).
29In Germany, where the public sector has the largest proportion of managers (Oesch, 2006, p. 112), women only represent a third of higher-level civil servants and a half of senior executives. Their situation is nevertheless more enviable than their private counterparts: women with similar educational qualifications are better paid in the public sector, while for men it is the private sector that is the better payer (Melly, 2005).
30In the United Kingdom, public sector employment is very feminized (65% women in 2004) in comparison to the private sector (41% women), which reflects the fact that there is much more part-time employment in this sector, 30% of employment compared to 24% in the private sector (ONS, 2005).
31In Sweden, 50% of all women work in the public sector, while the proportion of men is 23%, representing one of the biggest gaps among European countries (Oesch, 2006, p. 112). In Scandinavian countries, the development of the welfare state has provided inactive women and those working in declining industries with significant opportunities (Kohlberg and Kolstad, 1993, p. 60). For women, the public sector has therefore performed the role of a gateway to integration into the labour market, although in Sweden the salary advantage they gain from it is particularly modest (Gornick and Jacobs, 1998, p. 707).
32The second characteristic feature of public sector workers is that they are older than those in the private sector. This is particularly the case in France, where the proportion of those aged over 50 among agents in the public sector is 28%, while it is 18% in the private sector (Demke, 2004, p. 31). With the exception of the police and the armed forces where those aged 35 and below are in the majority, French public sector employees (civil servants or contract staff) are characterized by a demography that is largely the result of external recruitment waves and internal promotion, as well as the significant expansion of the local government civil service (Hugrée, 2011, pp. 69–70). This goes hand in hand with the longer length of service with the same employer in the public than in the private sector. The age gap is also perceptible in the United Kingdom, although in a less pronounced way: the proportion aged over 50 in 2004 was 28% in the public sector and 25% in the private sector and, more particularly, the proportion of those aged 35–49 was 44% against 36%. Similarly, employees in 2004 who had been in their job for ten years was greater, representing 40% of public sector employees against 28% of those in the private sector (ONS, 2005).
The persistence of a public service ethos
33Belonging to the public sector should not be reduced to the employment status and social characteristics of those who work in it. Employees also share a set of values that are in part those of the institutions to which they belong and through which they give meaning to their missions: neutrality, equality, and defence of the public interest. Sociological studies on the state have mostly shown that its authority is more effectively imposed when it is based not just on physical restraint but also on a symbolic legitimacy likely to produce adherence and consensus on the part of representatives of the state and on the part of users (Bourdieu, 2012). In Great Britain, various authors have sought to define what characterizes the public sector ethos: placing the public good ahead of personal interests, working with others collectively and anonymously, and resolving different problems with complete integrity (O’Toole, 1993). The debate that divides sociologists—especially British ones—is thus ascertaining whether these values have persisted beyond the upheavals of the last three decades.
34On the one hand, many sociologists argue that there is a specific character of norms of behaviour within public service (Du Gay, 2000, p. 12): they argue that despite the reforms that have shaken British administrative government throughout the 1980s and 1990s, certain values, such as accountability, honesty and impartiality, serving the public interest, and loyalty, can be thought of as a public service ethos that has been preserved over time (Pratchett and Wingfield, 1996). These values are not, however, common to all public services and in local services, agents can prioritize some more than others (Greenaway, 1995). Other authors highlight the dissonance between the discourse from rulers and civil servants on public sector values and the reality of practices on the ground after eighteen years of Conservative reforms (Richards and Smith, 2000, p. 47).
35On the other hand, analysing the effects of the reforms driven by Margaret Thatcher, sociological studies maintain that the spread of New Public Management has completely undermined the public service ethos: collective decision-making has been replaced by the duty of individual accountability, uniform salary scales by performance related pay, and the logic of cooperation by the spirit of competition (O’Toole, 1993, p. 5). Some think that the combination of public and private sector norms has encouraged the emergence of a new ethos specific to the—public and private—organisms involved in the new governance; the importation of private sector norms would turn agents away from a public sector ethos towards a public service ethos, including values such as pragmatism, procedural transparency, and a relationship with customers (Brereton and Temple, 1999, p. 471). From this perspective, we might question whether this sense of public service and the public interest cannot be found among some private sector workers, for example in the nuclear industry or in electricity generation companies (Linhart, 2009, p. 125). Others have also developed this idea by stressing instead the values that have different meanings in the public and the private sectors but that can be used in support of convergence between the two spheres, such as motivation, professionalism, trust, and impartiality (Plant, 2003). All these studies seem to suggest that the many reforms in public services have contributed to bringing the ethos specific to public sector agents closer to that of the private sector. However, there are few who have supported this observation with empirical data.
36The public service ethos should not be reduced to a set of values; it also reflects a subjective relationship to work. The vast majority of agents in public service have work satisfaction levels that are higher than their private sector counterparts (Baudelot et al., 2003); conversely they may declare themselves less satisfied with particular aspects of work, such as remuneration, autonomy, and promotion opportunities (Rainey and Bozeman, 2000, p. 459).
37In Britain, many more of them claim their jobs are “useful to society” or “enable them to help others,” a difference that has persisted, all other things being equal (John and Johnson, 2008, p 10). The public service ethos is thus defined by the desire to have a job that is socially useful and for those working in these positions to be satisfied by this. Despite alarmist discourse about the end of the spirit of public service, these differences in the ways of representing their work have widened after New Labour’s period in office (ibid., pp. 119–21). Public service agents place greater value on what benefits others and society, on the principles of accountability and integrity, and think less often of money and higher pay as the ultimate goals of work life (Rainey and Bozeman, 2000, p. 460). This greater indifference among public sector employees to remuneration is not however immutable. In France, it persists but is tending to fade: those working for the state are now more preoccupied by it and include it more directly in their job satisfaction, probably because of the objective deterioration of remuneration in this sector (Baudelot et al., 2014, pp. 144–9).
38The hypothesis of a public service ethos could also be applied to a much broader set of industrialized countries. Using responses to questions contained in the International Social Survey Program (ISSP) on approaches to work,  Pippa Norris shows that in European countries, particularly those in the former communist block, the idea of working in the public sector is associated with the idea of helping others. In contrast, people preferring to work in the private sector prioritize autonomy and independence more (Norris, 2003, p. 84). Ultimately, while working conditions in the public and private sectors have tended to converge, how workers represent themselves and engage in their missions remain specific to each sector. Private sector workers experience greater freedom at work, while those in the public sector are constrained by the need to be accountable, to meet transparent norms and ensure fairness among users. Public sector managers have less freedom than their private sector counterparts, but they have a greater feeling of accomplishment relating to their social roles and greater job security. All these differences combine to form a public sector ethos that goes beyond the specific configurations of each country.
Stronger trade unions in the public rather than the private sector
39Remaining in the occupational sphere, involvement in a trade union is also a significant point of difference between public and private sector employees. Indeed, the decision to join a union can meet certain political considerations, but it also primarily reflects a desire to be better protected at work, whether in terms of remuneration, work conditions. or more generally a desire to be defended in cases of conflicts with employers.
40The rate of trade union membership and the number of social conflicts have significantly declined in most OECD countries since the 1980s and 1990s, notably because of the rise in unemployment. There are various explanations for this disaffection. In the economic literature, the rate of union membership is studied from the perspective of the different advantages that union organizations offer depending on country and macroeconomic indicators, such as the unemployment and inflation rates or the level of benefits guaranteed to jobseekers (Checci and Visser, 2005). However, when the employment sector (public or private) variable is introduced, significant differences in the evolution of unionization among different European countries become apparent.
41Whilst it was private companies in the industrial sector that were the bastions of European trade unionism in the 1970s, it is now the public sector that has the highest rates of union membership, which has encouraged some authors to link the decline in trade unionism to the de-industrialization process (Scheuer, 2011, p. 66). Overall, the rate of union membership tends to be lower in the private sector, in small companies, and among workers with unstable jobs or working part-time (Ebbinghaus and Visser, 2000). In his study on Germany, Great Britain, Sweden, and Switzerland, Daniel Oesch shows that at the end of the 1990s, all things being equal, the likelihood of being a trade union member was much greater in the public than in the private sector (Oesch, 2006, pp. 173–8).
42However, this general observation can be explained in various ways depending on the country. In France, the rate of union membership has always been historically low and remains one of the lowest in Europe today. The various groups of affiliated trade unions represented 8% employees in 2004, with large disparities depending on employment sector; in the state civil service 1 in 6 employees is effectively a trade union member, while in the private sector is it one in twenty (Amossé and Pignoni, 2006, p. 409). The contrast with the Swedish case, where work relations are based on high levels of union membership, is “striking” (Anxo et al. 1998, p. 7): in 2007, around 70% of Swedish employees were union members (OECD, 2008) and, according to the European Trade Union Confederation, more than half of these were public sector agents. In the United Kingdom, the trade union presence in the public sector remains significant, and higher than in the private sector, although it is declining: the rate of union membership has fallen from 64.4% in 1993 to 58.4% in 2004 (Blanchflower, 2007, p. 4). In Germany, trade unionism has eroded considerably since the beginning of the 1990s, but it has been much more resilient in the public sector. Whilst private sector workers are less unionized than their public sector colleagues, and whilst this difference has tended to increase, this is in part because of the places of work are smaller and unions are less well established, in particular in the case of the service sector (Scheuer, 2011). Additionally, the social characteristics that distinguish unionized and non-unionized workers are not the same in the two sectors. In the United Kingdom, for example, having a high level of education and a higher qualification increases the likelihood of a person in the public sector being a union member, while in the private sector these characteristics are shared more by the non-unionized (Blanchflower, 2007, p. 8).
Public–private: different world views?
43While American sociology and political science can draw on comprehensive quantitative studies showing the differences between public and private sector employees (Brewer, 2003), European studies have been more targeted on either countries or on a certain type of practice. Some such studies highlight different behaviour in the two sectors, in areas as varied as cultural practices, family relations, and educational strategies. The seemingly fragmented nature of these sets of themes in reality reflects a dimension of the same relationship to the world that we might term lifestyle (Bourdieu, 1979, p. 56). If we also consider the effects of the public–private divide on political orientation, we might pose the question of whether the differences between employees in the two sectors actually extend far beyond the rigid framework of employment conditions.
Lifestyles and cultural practices
44The differences between public and private sector employees are not limited to the occupational sphere, but there are few studies that address them exhaustively: they are normally empirically anchored nationally and only address a single dimension, which lends the notion of lifestyle a heterogeneous character. Most often, nevertheless, they investigate the same group: the middle class.
45On cultural practices, various British sociologists have highlighted the working class’s relative lack of investment in culture, in contrast to the very strong commitment of the middle class. Within the latter, public servants and teachers are more “omnivorous” than private sector managers and closer to legitimate culture (Savage et al., 2009, p. 181). Some propose exploring the differences within the middle class by identifying lifestyles (Devine et al., 2005). In particular, they observe consumption and leisure practices and identify three factions within the middle class: a group whose cultural and leisure practices are ascetic and mostly correspond to public sector professionals (teachers, doctors, and welfare state employees), a so-called “post-modern” group made up of private sector professionals and an “undifferentiated” group, in other words one whose cultural commitments are weaker and less distinctive, grouping mostly private and public sector managers (Savage et al., 1992, pp. 108–10). Whilst such a breakdown may not entirely correspond to a public–private divide, the issue of public sector membership nonetheless runs through it. The closeness of a large proportion of professionals in the public sector is a result of the fact that those who belong to it rely principally on their cultural capital, capital that is largely dependent on the state that, through the education system, ensures its reproduction and its value. This results in a new division between, on the one side, the public sector middle class—made up of professionals and increasingly feminized— and on the other, the private sector middle class, most often propertied and entrepreneurial (ibid., p. 218). This conclusion can also be linked to Edmond Préteceille’s analysis of the phenomena of urban segregation resulting from residential strategies: according to Préteceille, corporate executives tend to be concentrated in areas occupied by the economic and financial elites, while those in the public sector remain in more socially mixed areas (Préteceille, 2006, p. 85), at the same time preserving the social selectivity of school places.
46From the standpoint of family relations, some studies show that working in the public sector encourages a better work/family life balance. In Sweden, for example, a father working in the public sector tends, all things being equal, to use much more of his parental leave than his private sector counterparts (Bygren and Duvander, 2006). With the public sector having more flexible working hours and less hierarchical pressure, employees working in it are less exposed to conflicts between family and work overall (Scherer and Steiber, 2007).
47Finally, belonging to the public or private sector conditions certain practices and certain choices in children’s education. Based on a longitudinal study of the trajectories of children born in 1971, British researchers have shown that public sector middle class parents are much more likely to send their children to state schools than those working in the private sector; moreover passing through one or other system (state or private) has a durable effect on representations, dispositions, and thus, ultimately, on the trajectories of the pupils concerned (Power et al., 2003). This form of loyalty to the public sector had previously been revealed in the case of France: an education spent entirely in state schools significantly increased the likelihood of doing a job in the public sector subsequently: for men educated in the state system, the probability of working in the public sector is 44% while it is 25.5% for those coming from the private system (Langouet and Léger, 1997, p. 114).
48In terms of political orientation, the division between public and private sector workers is very notable, even though the tools for measuring this may invite caution because these are mostly opinion polls. Traditionally, the votes of public employees go more to parties that support the maintenance of the welfare state and the defence of access to public goods,  while those in the private sector are more sensitive to the laws of the market and more exposed to the effects of international competition (Kitschelt, 1994).
49In France, the large number of public employees (a quarter to a third of electors) lends this group a particularly significant political weight, regularly highlighted by political scientists (Boy and Mayer, 1997; Rouban, 2005). According to the 2004–2005 European Social Survey, it is in France that the public–private divide is the most significant in terms of political orientation: a public agent is 1.6 times more likely than a private sector worker to claim to have voted on the left at the last election; this ratio is also significant in Germany (1.3) but less so in Sweden and the United Kingdom (1.1) (Mayer, 2010, p. 114).
50In Sweden the divide between the public and private sectors has been much analysed from the perspective of political values (Hoel and Knutsen, 1989, p. 198). Public sector employees claim to be much more significantly attached to the welfare state than those in the private sector, although this difference has diminished (Svallfors, 1995, p. 68). Based on a longitudinal study of voting by social class, employment sector and gender between 1970 and 1995, Oddbjørn Knutsen (2001, p. 313) emphasized that the opposition between public and private employees encompasses several dimensions: divergent economic interests, distinct cultural environments, and different work experiences. In the occupational sphere alone, the service relationship many public sector workers provide exposes them to social problems that they have to respond to by drawing on public service values. These values also partly overlap with those they were taught throughout their education. A comparison of surveys conducted in 1970 and 1995 in Denmark, Sweden and Norway, leads the author to suggest that the impact of social class on voting is declining while that of gender and employment sector is increasing (Knutsen, 2001, p. 345).
51In Britain, differences in the political orientations between public and private sector agents have also been widely analysed. Patrick Dunleavy’s (1980) work has shown that the working class can be distributed across two axes, depending on whether they are public or private sector employees and depending on whether they are union members: Labour is stronger among the most unionized public sectors and the Conservatives have stronger roots among less unionized groups and the most fiscally constrained. This division is found in the political orientations of members of the middle class who have a different relationship to the state. An analysis of electoral polls in the 1980s shows that professionals, mostly associated with the public sector, have a higher tendency to vote for the Labour Party, while managers opt more for the Conservatives (Savage et al., 1992, pp. 194–7; Butler and Savage, 1995, p. 286). This political differentiation involves two different relationships to the state: the professionals, often paid by and financially dependent on the state, have an interest in supporting the maintenance of a certain level of intervention by the latter in society, while the managers, often employed in the private sector, are less inclined towards this, since they benefit from outsourcing and privatization of public services. This finding is confirmed by a more recent study from the British Household Panel Study in 1997: public sector workers demonstrate and vote against the Conservatives, while those in the private sector provide the support for the Conservatives’ reforms, but this division is modulated depending on each group’s social composition and place of residence (Dorling et al., 2001, p. 65). Similarly, white-collar public sector workers have been the staunchest defenders of a public health service and state education, encouraging some to deduce from this a new political divide between, on one side, working-class public sector users alongside public sector managers and professionals and, on the other side, the private sector middle class (Roberts, 2001).
52The link between employment sector and political orientation can therefore take on different forms depending on each country. In a study of eight industrialized countries  and based on data provided by the Eurobarometer, Knutsen (2005) extends his reflection on the specificity of public sector agents in terms of political choices. Employment sector—public or private—has a significant but modest impact in each of the countries studied: the correlation is greatest in Denmark, representative of a social democratic state, and weakest in Ireland, the incarnation of a social liberal state (Knutsen, 2005, p. 604). Based on another source, the 2002 European Social Survey on 22 European and extra-European countries, Luc Rouban (2005) also identifies various significant results regarding the impact of employment sector on the sociopolitical cultures of employees in Europe. Firstly, public sector employees vote more often than those in the private sector (their participation rates are 85% against 74% for those in the private sector); they are also on average more often members of political parties, in particular in Denmark and Sweden (it is moreover French civil servants, who often have a reputation for activism, who— along with the Spanish—are the most reticent to engage politically). Even in the United Kingdom, where the public sector has long been affected by liberal reforms, 54% of public sector employees declare themselves to be close to a political party (against 50% in the private sector) (Rouban, 2005, p. 62). In a more recent study, also based on the European Social Survey, Markus Tepe (2012, p. 253) refines the observations previously established by introducing distinctions between sectors: central government agents vote in same way as those in the private sector with the same level of education, while those in the health and education sectors retain a loyalty to the left. However the latter may move closer to the right as they move up the hierarchy.
The habitus of the public sector middle class
53Many empirical and theoretical studies have thus shown how the specificity of functions assigned to public sector agents encourages them to share certain values, in particular as a result of their exposure to social problems. Most often, these common features are rallied behind the notion of a public service ethos that is sometimes conceived of as a simple consequence of the work and sometimes an effect of the sector (Davies, 2012, pp. 29–30). From the panorama that we have established from the literature, it seems to us that the idea of a pubic sector ethos can be amended on several points.
54Firstly, the public service ethos, defined as a set of values shared by employees working in the service of the state, has different meanings depending on the place occupied in the social hierarchy. Moreover, the values that define the public service ethos are not fixed once and for all. The introduction and then spread of New Public Management norms made these values an issue in the clashes between agents themselves (McDonough and Polzer, 2012). For agents who continue to identify with public service missions, these values extend far beyond the context of the occupational sphere alone: in these cases, the public–private divide has repercussions for lifestyles, cultural practices, and political orientations, which encourages us instead to use the term habitus, which is broader and more encompassing than that of ethos. The occupational socialization within the public sector has repercussions that extend far beyond simply the conception of the job. In a study of public hospital doctors faced with the introduction of private partners, a British sociologist was able to show that the modification of work structure in a more managerial direction only has a small effect on the doctors’ representations and beliefs. They take these imperatives into account while continuing to develop practices that conform to their habitus: they support a conception of public service that, far from being individual resistance, reflects a common position among this group of actors who work with patients (McDonald, 2009).
55If there really is a set of representations and dispositions incorporated and shared by workers in the public sector beyond their work lives, these should be understood in light of the position of these employees in society. In the recent period, senior managers have experienced significant changes in terms of their roles and status: their rights and protections have been reduced, while in exchange their salaries have increased. Thus, in the United Kingdom, around 40% of management positions are open to competition from executives coming from the private sector (Demke, 2004, p. 150). Conversely, in some cases, they can be dismissed if they have not met their objectives. This blurring of the boundary between the public and private sectors concerning the recruitment of senior managers has an effect on work cultures and more generally on representations. As Janette Webb (1999) has clearly shown, there is a growing divide between entrepreneurial strategists and welfare professionals: the former, most often men, adopt business values and identify with them, while the others, mostly women, favour caring values (Webb, 1999, p. 763). While they all seem to belong to the same world, the boundary separating them is in fact a very marked one, including the representations they have of each other: senior managers think of welfare professionals as pure idealists, while the latter see the former as the representatives of power (ibid., p. 763). This growing divide between the “right hand” and the “left hand” of the state (Bourdieu, 1993) is confirmed by other studies that seek to show the convergence, in terms of values and conception of functions, between public and private sector managers (Rainey and Bozeman, 2000; Buelens and Van den Broeck, 2007). Moreover, sociologists who maintain there is a phenomenon of hybridization of values between the public and private sectors often rely on interviews with executives who are quite senior within their organizations. Thus, Michael Brereton and Michael Temple (1999, p. 464) cite the words of a director of a regional health authority who considered that relationships with private sector partners were “stimulating.” In the 1990s, British sociologists even highlighted the effect of state reforms on the changes in the state middle class. They showed that these reforms (decentralization, managerialization, increased supervision, employment at lower levels made more flexible, etc.) favoured top managers who are given more responsibilities, who are better paid (bonuses) and who acquire new skills. The re-unionization of public sector employees and the growing social conflicts within administrations illustrate the internal fragmentation process among public sector employees (Carter and Fairbrother, 1995). In light of these studies, we conclude that New Public Management, in Britain in particular, exposes socio-occupational groups making up the public sector to centripetal forces that bring together public sector middle class dispositions and representations on the one hand; and, on the other, to centrifugal forces conducive to a convergence between public and private sector senior executives.
56The hypothesis of a habitus common to the public sector middle class is particularly emphasized by studies devoted to front line workers: that is to those—middle managers or junior agents—who are in contact with the public (McDonough, 2006; Serre, 2009, p. 300). Most of them share a universal vision of the world that places public interest above private interests. Through their daily practices, they embody and reproduce this conception of the public good in phase with the objective structures of public service, thus rendering it as “self-evident”. Of course, this vision is never established once and for all: it is the subject of struggles and called into question, especially since the New Public Management reforms. In this respect, the question of belonging to an employment sector (public–private) should today be linked to the decline or persistence of a representation of society in terms of social class. In France, socio-occupational category and the public sector are clearly discriminating factors in declarations of feelings of affiliation to a social class and in the subjective positioning within the middle class (Pélage and Poullaouec, 2007). Beyond the French case, these results also constitute original groundwork for renewed debate on the symbolic effects of belonging to the public sector in Europe.
57* * *
58The divide between public and private sector workers is not a substitute for an analysis in terms of social class, the outlines of which remain to be defined at the European level. The fact remains that, within the same social group, belonging to the state sphere is a strong marker of social practices and visions of the world. Although the divide between the public and private sectors rarely appears in statistical and comparative studies at the European scale, it is still a significant dimension at the national scale for the set of countries studied (Germany, France, Great Britain and Sweden). Following the reforms introduced in most European countries in the name of New Public Management, many studies have shown a convergence between public and private sector norms from the point of view of work conditions. However, despite successive reforms and different legal realities governing the notion of the public sector in Europe, working in the service of the collectivity still produces a set of characteristics today. Belonging to the public sector should thus not be reduced to the set of working conditions and hierarchical relationships current within it. It also implies a specific relationship to the state, the public interest, and also to public affairs, perceptible in cultural, trade union, and political practices. This correlation is far from universal and is fading at the highest levels in the hierarchy as reforms of states spread.
59This bibliographic overview suggests we should not too hastily abandon inclusion of this divide within what could be a European socioeconomic classification. Within the context of the ESeC (European Socio-economic Classification) that was undertaken between 2004 and 2006 to design this sort of statistical tool, this question was raised by French studies, and more generally when the social structures of countries in Southern Europe had to be described (Rose and Harrison, 2010). However, in continuing this work centred on the comparison of the possibility of social mobility in different countries (Erikson and Goldthorpe, 1992), European classification projects have not made the difference between public and private sector workers visible (Rose, Pevalin and Elias, 2001; Rose and Harrison, 2006). Thus, there is no consensus regarding its inclusion in the prototype ESeC constituent variables among the experts charged with deciding these (Brousse, 2008; Penissat and Rowell, 2012). Moreover, in the most recent project led by INSEE, entitled ESeG, the willingness to account for the public–private divide came up against the lack of harmonized variables at the European scale that could quantify it. Yet, the two sectors are still clearly different: public sector employees prove far more feminized and better qualified in all European countries. The review of various studies devoted to public sector employees invites us to speculate on the “describability,” in particular statistically, of this group, despite the significant differences in national histories from the point of view of politics and employment. The issue of public-private distinction appears now to be sufficiently documented and shared for it to be central to the sociological understanding of the construction of salaried workers across Europe. Indeed it is precisely through this type of approach that we avoid reducing the European workforce to an axiom, or reifying it as a bureaucratic artefact.
The expression “middle class” in English does not correspond to that of the “classes moyennes” in France, but encompasses the dominant classes and the upper portion of French intermediate occupations (professions intermédiares) (Bidou-Zachariasen, 2000).
They are mostly civil servants and few are public salaried employees but there is no official statistic available providing a distribution in terms of employment status.
Using a similar definition, the difference between statistics from government sources and from surveys is around 670,000 employees (ONS, 2005).
The study focuses on Italy, France, and Great Britain.
This is a survey of a wide range of countries including Bangladesh, Bulgaria, Canada, Cyprus, Czech Republic, Denmark, France, Germany, Great Britain, Hungary, Israel, Italy, Japan, the Netherlands, New Zealand, Norway, Philippines, Poland, Portugal, Russia, Slovenia, Spain, Sweden, Switzerland and the United States of America.
In Norway, public sector employees are even seen as swing voters for public sector reforms (Rattsø and Sørensen, 2004).
Germany, Belgium, Denmark, France, Great Britain, Ireland, Italy and the Netherlands.