The Blackwell Companion to Sociology (61 page)

attempts to root political change in the concept of class, and is especially

influential in developing a gender-focused approach to politics. Skocpol is per-

haps the best known publicist of a much deeper movement within political

sociology which doubts the value of rooting political mobilization in economic

and social interests, and which has been termed à`new institutionalism.''

By the later 1990s it was clear that political sociologists had largely

abandoned the idea of rooting politics in social bases. For Boli and Thomas

(1997) and Ramirez et al. (1997), the globalization of politics entails the declining significance of nation states and by implication the classic arena in which

social forces mobilized. For other writers, such as Korpi and Palme (1998) and

Gornick and Jacobs (1998), it is the structure of the nation-state that determines social processes and not the other way around. Korpi and Palme's paper is a

particularly interesting example of how the institutionalist approach undermines the classic paradigm. They argue that political movements determined to reduce

inequality are actually counterproductive and end up by increasing inequality.

Thus, there is no social logic by which progressive social forces can effectively implement change through social action.

A second trend toward the dissolution of thè`social'' can be found in the work

of social historians. During the 1960s and 1970s, social historians, following the lead of Edward Thompson, Herbert Gutman, and Barrington Moore, championed a mode of class analysis which examined how class mediated between

social structure and political mobilization. Intellectual currents in the 1980s

indicated a growing unease with this approach. In a well known study of

Chartism, Gareth Stedman Jones (1983) emphasized the need to examine care-

fully ``languages of class,'' that is to say the precise way that class was operationalized in language. Jones's arguments had an explosive effect in generating

what has been termed thè`linguistic turn.'' This current, associated particularly with the historians Patrick Joyce (1990, 1993, 1994), James Vernon (1994,

1993), and Bill Reddy (1987), situated political movements in terms of the

linguistic categories available to people in different cultures and at various

times and also suggested that the causes of class inequality were political rather than economic.

A third important development during the 1980s was the rise of a new kind of

systems-based state theory. This was indebted to the structuralist currents of the 258

Mike Savage

1970s, notably those associated with Louis Althusser and Nicos Poulantzas (see

Jessop, 1983), and during the 1980s this matured as a distinct Regulation School (notably Aglietta, 1979; Boyer, 1990; Jessop, 1995). Writers within this school

developed powerful sociological interpretations of political change, with Jessop et al.'s (1988) analysis of Thatcherism being especially well known. The key

features of this approach lay in a different understanding of the relationship

between society and politics to that found in thè`social bases'' approach. Rather than relating political movements or forces to specific social groups, the emphasis was on showing the interrelationship between changing modes of political

regulation and the dynamics of the capital accumulation process. For Jessop et

al. (1988), the decline of social democratic politics and the rise of Thatcherism could be seen as related to the transformation from Fordist to post-Fordist

politics, and an attempt to establish a new kind of political hegemony.

The foregoing discussion suggests that the crisis of political sociology is closely related to the erosion of the idea of thè`social'' itself. In different disciplines and for varying theoretical reasons, thè`social'' is elided, and the autonomy of

politics is reasserted. Even attempts such as those of cultural historians

or political economists have tended to follow the same road (sometimes unwit-

ting) to institutionalism. All these approaches have provided valuable, though

contradictory, ways of reflecting on social and political change, but all of them have moved away from the classic paradigm of political sociology. In the next

section I discuss how the defense of the classic paradigm has become attached to debates about the relevance of social class, and evaluate the lessons of recent

research for considering whether the classic paradigm can be salvaged.

The Working Class and Politics: the Linchpin of thè`Classic'' Paradigm

In the classic paradigm class became central to political sociology because of its ability to offer a way of linking society and politics. Classes were rooted in social relationships but had political expression. Thus, when Sartori reflected on the

nature of political sociology in the late 1960s, it was not accidental that his

remarks were almost exclusively concerned with the relevance of class. Needless

to say, there were very different ways of examining class, with some writers

anchoring political processes more firmly in class structures than others did.

What is important for our discussion is the recognition that the crisis of political sociology is also the crisis of class analysis.

It is within British research that the defense of class-based approaches to

political alignments has been most trenchant, and I therefore focus on it here.

In the mid-1960s, class-centered analyses of politics clearly held sway, expressed most memorably by Pulzer (1967) in his phrase that ``class is the basis of British politics: all else is embellishment and detail.'' This argument was backed up by Butler and Stokes's (1969) study ± the first using the British Election Surveys, which were first commissioned in 1964 ± which underscored the centrality of

class. From the 1970s, the consensus that class was fundamental to political

Political Sociology

259

alignments has faded, and during the 1980s a bitter dispute broke out concern-

ing the idea of class dealignment. This debate has been instructive, since it clearly pitted sociologists against political scientists, and thereby revealed key differences in the methodology of political sociologists and mainstream political

scientists. The rallying cry was that of class, especially the politics of the working class, but the debate on class dealignment also involved complex methodological

issues about how to understand the relationship between class and politics, as

well as substantive arguments about how social change was eroding the relation-

ship between class and politics. In the early 1980s in the UK, attention focused on the shifting politics of the working class rather than that of the middle class.

For political scientists such as Crewe and Sarlvik (1983), Kavanagh (1990a), and Crewe et al. (1991), the declining propensity of manual workers to vote Labour,

and the general hegemony of Conservative politics during the 1980s, appeared to

sunder the traditional association between class and politics.

Political sociologists defended class analysis in a number of sophisticated

ways. One of them was by invoking a more rigorous definition of class structure.

In place of the crude division between manual and non-manual workers that had

dominated in older literature (Alford, 1967), Heath et al. (1985, 1991, 1994)

and Marshall et al. (1988) emphasized that if the manual working class and non-

manual middle class arè`decomposed,'' it is possible to demonstrate that class

and voting continue to be associated. Thus, in 1983, the middle class could be

divided into small employers, who voted 71 percent Conservative, and routine

non-manual workers, who voted 53 percent Conservative.

One innovation was to examine class politics in terms of relative class votes

rather than absolute numbers. To illustrate, although Labour voting by manual

workers declined between 1966 and 1983, support among middle-class voters

fell far more, indicating, according to Heath and his colleagues (1985, 1991,

1994), that there was no general trend toward dealignment. However, on the

basis of recent studies there are general questions raised about this thesis of

``trendless fluctuation'' (see Weakliem, 1995; Evans et al., 1999), and the 1997

election marked a further striking drop in class voting.

Furthermore, to the extent that relative class voting does persist, this cannot

be taken as evidence for the importance of class cultures in political alignments (Heath et al., 1985). Crewe and Sarlvik (1983) and Franklin (1985, 1992)

emphasize that the decline of class-based solidarities is linked to the tendency for voters take on à`consumerist'' and ``reflexive'' approach to politics (see also Dunleavy and Husbands, 1985). Crewe and Thompson (1999) show that there

has been a remorseless decline in strong party identification over time (in 1966, 44 percent of the electorate were strong supporters of a political party, a figure which dropped to 16 percent by 1997). Bartle (1998) has also shown that social

class is not an especially important determinant of people's party identification or political outlook.

The debate on class dealignment has had relatively little international impact.

In the USA, studies of voting and political alignments have over a number of

years shown relatively little interest in class (for example, Campbell, 1960).

However, some intriguing studies of American voting and politics which draw

260

Mike Savage

on British approaches have appeared in recent years (Brooks and Manza, 1997a,

b). Brooks and Manza argue that relative odds of members of one class rather

than another voting for one president rather than another did not change

significantly between 1960 and 1992. They argue that the relative importance

of gender and race increases, and class is at the very least stable. However,

although they show that there are significant class effects, these turn out to be rather unusual ones. Professionals support Democrats, while managers support

the Republicans, indicating that it is political division within the middle classes, rather than between working and middle class, which is significant to political

alignment. There is some British research pointing in the same direction (Savage et al., 1992). Some professional groups (especially teachers, social workers, and academics) lean more to the left than do private sector managers, members of

security services, and large employers (see Heath et al., 1991; Heath and Savage, 1995). Furthermore, Heath and Savage (1995) show that there is some evidence

that professionals have been shifting their allegiance away from the Conservat-

ives since the 1970s.

In concluding this section I finish with a general point and a rider. The general point is that even the most stubborn defenders of class analysis have had to

modify their arguments so much in the face of recent political change that it is now implausible to claim that a clear class-based approach to politics remains

intact. But my rider is this. As Brooks and Manza (1997b) show, and as is also

clear in the British case, there is evidence that in place of the traditional divide between working and middle class, fractures within the middle classes may be

becoming more politically important (see also Savage, 2000). In the 1997 British general election, Labour won back the level of support from the working class

which it enjoyed in the 1960s, but also won unprecedented support from the

middle classes. Indeed, the shift of middle-class support to Labour in 1997 was

far larger than the shift of the working classes to the Tories in the 1980s. This point is interesting because it suggests that if we accept that the traditional social base approach to political sociology is dead, there is at least the possibility that one way of salvaging elements from it is to consider not how political alignments between working class and middle class energize the political field, but how

relationships within the middle classes now play a central role. This is the idea I wish to develop in the remainder of this chapter, though I want to emphasize that taking this idea seriously means rethinking a new form of class analysis (see

Savage, 2000).

New Directions in Political Sociology

An interesting way of re-energizing the agenda of political sociology is to focus not on the social bases of political alignments, but on the sociology of thè`political field'' itself. The crucial difference between these two perspectives is that whereas the prime interest of the former is the sociological determinants of political alignments, the latter focuses on the relationship between the legitimate sphere of organized politics as a whole and other kinds of social relationships.

Political Sociology

261

Instead of examining the determinants of voting for specific parties, people or

causes, it examines how the social relations and networks involved with legit-

imate political activities (voting, organizing, campaigning, etc.) intersect with those of other social practices. This offers the prospect of a new kind of political sociology which examines the social construction of the boundaries between thè`political'' and thè`social,'' and is attuned to divisions not within the political, but between the social and political.

This is a complex argument which I can hardly do justice to here. A starting

point is to emphasize the extent to which formal political activity has been

removed from everyday life. A good example is in voting. Undoubtedly, in liberal democracies, voting is the most important way in which the majority of people

interface with the world of formal political activity. Nonetheless, voting is a