Political Science and Social Conflict

In Marxism the struggle between the classes over the disposal of the surplus yielded by the production process is an inescapable feature of all class societies. Marxism's attachment to the working class has nothing to do with nostalgia for the "simpler" politics of the past or anything like that. Instead it is based on the observation that the (subordinate) class relationships an absolute majority of the world's population has to enter into to make a living provide grounds for drawing together the largest numbers possible to pursue socialist politics. i.e. Fighting and building a society free of the class relationships that have characterised history since the dawn of civilisation. Marxist politics therefore has concentrated on building political parties and movements that articulate the interests of workers under capitalism and the trajectory these interests have in moving beyond capitalism.

Whereas class and class conflict is absolutely central to Marxism, this is not the case for political science as a discipline. In this piece (recycled from a presentation I gave about five years ago), I look at the place social conflict, (or to use the preferred term, 'cleavage') occupies in political science and how this position has tried to make sense of the new politics that have come with the restructuring of capitalism that has taken place these last 30 years.

In their seminal 1967 paper, ' Cleavage Structures, Party Systems, and Voter Alignments', Seymour Martin Lipset and Stein Rokkan argued that political parties arose in response to and were conditioned by four cleavage structures: the conflict between centre/periphery, state/church, land/industry, and owner/worker. Each of these developed in that order, with the outcome of each effecting the development of subsequent cleavages. The development of these cleavages predated the advent of mass suffrage in Western Europe, meaning that when mass parties did emerge they expressed the various interests rooted in cleavage structures, and were dependent on them for the necessary resources.

Lipset and Rokkan argued that the development of cleavages effectively ended with the employer/worker contradiction. The 1920s-1960s period saw parties more or less reflect existing cleavage structures: that is despite the experience of fascism and war and the dislocations these caused continental party systems, the overall political trend favoured stability: what were the regional, rural, religious and class parties of the 1960s tended to be the same parties 40 years previously. Lipset and Rokkan also suggested that a comparative analysis of parties emerging during the wave of mass suffrage were far more likely to be durable than newer parties built after this wave had receded. As far as they were concerned the cleavages bequeathed by the early development of capitalism had built West European party systems, and were stabilised by the persistence of the four cleavages.

Richard Rose's and Derek Urwin's 1970 paper, 'Persistence and Change in Western European Party Systems Since 1945' set out to verify Lipset and Rokkan's observations by looking at the electoral support of West European parties from 1945 to 1969. If there were no significant shifts in party’s electoral support, then Lipset and Rokkan's stabilisation hypothesis was confirmed. Rose and Urwin concluded that the data showed “the electoral strength of most parties in Western nations since the war had changed very little from election to election, from decade to decade, or within the lifespan of a generation” (1970, p.295).

Maria Maguire's 1983 paper, 'Is There Still Persistence? Electoral Change in Western Europe' used a similar methodology to update Rose and Urwin's results. She looked at the electoral performance of 84 parties (using an inclusion criteria specifying that qualifying parties must have fought at least three elections across her time periods and polled 5% or more on at least one occasion) from 15 West European countries (organised into 'Britain and Ireland', 'Scandinavian', and 'Continental' groupings). The two periods selected by Maguire were 1948-79 to provide a broad overview comparable with Rose and Urwin, and the 1960-79 short period looking at the degree of changes in greater depth. The short period study analysed 80 parties, arrived at using the 1948-79 inclusion criteria.

Maguire subjected produced data using three tests. The first test defined significant trends as swings in support of 0.25% or more either way per year. For Rose and Urwin 70% of parties experienced a ‘nil’ trend, whereas on Maguire’s ‘long’ scale this figure dropped to 64%. This change becomes more substantial on the short measure; here the nil trend only encompassed 51% of the 80 parties, 21 of which experienced cumulative change of six per cent or more over the 19 years.

Looking at electoral fluctuations (scores here computed by the difference between a party’s largest and least share of the vote), median elasticity on Rose and Urwin’s measure was 7.0%. On Maguire’s long measure this had risen to 7.8%. Some 49% of the parties experienced a median score of 8.0% and over, and 29% of the total sample were in excess of 12.0%. Surprisingly, on the short measure median elasticity fell to 6.7%, - parties experiencing fluctuations in excess of 8.0% fell to 38%. For Maguire the problem here is if fluctuation is measured with reference to extremities of range, accuracy may be compromised by atypical election results. Instead Maguire computed another set of figures based on a “normal” range of fluctuation, based on the standard deviation of the mean vote. On her long measure the ‘median party’ had an SD of 2.6, with 56% of parties having an SD lower than 3.0. Reasonably similar figures were obtained on the short measure, with a median party SD of 2.7 and 61% of parties experiencing an SD lower than 3.0. This indicates that any trends occurring were not manifesting themselves via high rates of vote fluctuations.

Maguire then performs a measure of electoral volatility. This measures the net change within ‘electoral party systems’ by looking at individual vote transfers. Used on individual parties, Maguire calculates the percentage change in a party’s vote between each pair of elections over her two timescales. On the long period average volatility ranged 0.4% - 11.8%, and on the short it was 0.4% - 8.3%. In the long 72% of parties had a volatility below 3.0%, while in the short it was 65%.

Cross tabulating volatility with trends data (with scores organised into low = 0 - 1.5%, medium = 1.6 - 3.0%, and high = 3.0% +) shows that on the long period 45% of ‘nil trend’ parties had low volatility while 18% had high, while those registering significant trends (whether decline or rise in support) were 40% high and 10% low. In the short period 55% of ‘nils’ experienced low and 13% high volatility, whereas ‘significants’ were 54% high and 3% low. Thus while high fluctuations were a minority in nil trend parties, their status as ‘nil’ did obscure the fluctuations that did take place.

Finally a more accurate picture of volatility must take party size into account. Maguire does this by expressing average volatility as a percentage of a party’s average vote, with results grouped under low (10% or less), medium (11-20%), or high (20%+). In both Maguire’s periods 30% of parties fall into the low category, with more ‘nil’ parties experiencing high volatility (48% and 40% on the long and short measures respectively) than those undergoing significant trends (27% and 38%).

What role did party characteristics play in unfreezing the cleavage foundations of party systems? Looking at the parties, those originating in the ‘interwar’ (1914-39) and ‘new’ (1939+) periods experienced less instability than pre-1914 parties (only 50% of ‘old’ parties registered a nil trend, as compared with 73% and 70% respectively for the others, with 39%, 10%, and 19% experiencing significant downward trends). In the short period the respective figures were 54%, 48%, and 43% experienced significant trends, with 43%, 21%, and 26% of the total sample in receipt of downward trends indicating the 1960-79 period was very unstable. Maguire also looks at party family and electoral strength to see if they have any bearing on trends.

Overall Maguire concludes that while there is no apparent relationship between electoral fluctuations and trends, electoral change within party systems appears long term and gradual. The freezing of cleavages described by Lipset and Rokkan appeared to be undergoing a slow thaw, but why?

New Cleavages?
In their 1987 book, Elections and Voters, Martin Harrop and William Miller suggest additional cleavages may account for parties’ growing experience of significant change. These are grouped under two headings:

1) The impact of the state: the growth of the (welfare) state divides employees along public/private sector lines, consumers between the consumption of public and/or private services (typically transport and housing), and between net recipients of welfare benefits and net contributors. Harrop and Miller speculate that those tied to public services in some way are politically predisposed toward non-conservative values, or in the case of those who depend on the state for income, apathy and disengagement from politics.

2) Post-industrial influences: if we accept the thesis that manufactured “material” commodities are increasingly displaced by knowledge and service production; in conjunction with growing affluence these structural transformations will have significant impacts on the electoral performances of parties. Goldthorpe et al's seminal 1969 study of growing affluence among workers investigated whether these changes were affecting working class self-identification, and whether this impacted on the traditional working class attachment to the Labour party. Abramson and Inglehart (1995) tested whether there was a relationship between economic security and the adoption of postmaterialist values, finding that this was indeed the case.

There are other approaches that look more to the changing characteristics of existing cleavages rather than hypothesising the existence of new ones. Colin Crouch's pessimistic exploration, Post-Democracy (2004) argues recent changes in the class structure have fed into “post-democratic” developments. Identifying the manual working class as the working class per se and opposing it to the “middle mass” of “diverse and heterogeneous groups of professionals, administrators, clerical and sales workers, employees of financial institutions, of public bureaucracies, and of welfare state organizations” (p.57) Crouch is able to bring out important sectional divisions within this differentiated non-manual strata: the horizontal division into the public and private sectors, and the vertical split between senior managers and professionals who are structurally closer to capital, and routine workers likely to experience proletarian conditions of employment. The political amorphousness of the subordinate levels has allowed politicians to identify “lower middle class” concerns with business interests, and “encouraged to seek no means of improvement other than for themselves” (p.60). In terms of political impact, if this is the key constituency campaign managers are aiming for it means adopting more of a catch-all kind of politics.

However for Crouch this group represents a difficult problem for democrats – they are most susceptible to manipulation owing to the absence of an autonomous political profile and tend to be the most politically passive of groups, whilst being the most rapidly expanding sector in Western societies. Mainstream parties however have failed to acknowledge let alone address the real conditions of its everyday existence. For example, contemporary politics have shown a remarkable inability to grasp how downsizing, growing workloads, and mounting pressures on free time are common experiences of the workplace, meaning discontent around these issues could escape ‘official’ political frames. Such grievances could potentially lead to progressive politics animated by the “objective political agenda”, or find a populist rightwing expression - as the emergence of formations like the Lijst Pim Fortuyn, UKIP, and Veritas suggest.

Therefore, in sum there is a long term tendency toward the fragmentation of political systems. But parties themselves are not passive actors. In an earlier period mass parties, particularly those of the left, served to integrate people into political systems that, prior to the achievement of mass suffrage, excluded them. Is it possible for parties to perform a similar role again and counter act disintegration?