Class identification in New Zealand: An analysis of the relationship between class position and subjective social location

Abstract

Using data collected by two New Zealand surveys in 1999 and 2009, I explore the connection between the objective social class positions of individuals and their own subjective perceptions of these circumstances. Class position is ‘operationalized’ using a newer variant of Goldthorpe’s schema, the European Socioeconomic Classification (ESeC). Through regression analyses, it is demonstrated that ‘objective’ forces contain positive predictive consequences for self-placement. More importantly, the results suggest that as predictors of subjective class, the effects of class have endured while those of education and income – understood here to represent measures of socioeconomic position – have declined. The empirical evidence produced suggests that class continues to generate subjectively salient identities, leading one to deduce that there are no grounds for stating that it is no longer a significant feature in society.

Keywords

class analysis class identification New Zealand quantitative analysis subjectivities

The recent global financial crisis combined with the continuing growth in income inequality in many countries have spurred an interest in topics connected with class (Atkinson, 2010a; Evans and Tilley, 2012; Weeden and Grusky, 2012). The BBC’s 2011 Great British Class Survey, which represents the largest survey ever on social class in the UK, is testament to the resurgence of class in mainstream discussions (Savage et al., 2013). Many citizens have engaged in political movements which question the capitalist mode of production, revealing how a select few have immensely benefited from the crisis as some have slipped further into poverty. The resurgence of class politics speaks to an earlier theory developed by Weber (2000 [1914]) that during periods of economic transformation, ‘class situation’ is pushed to the foreground as stratification based upon status recedes.

In the early 1980s, New Zealand engaged in a neoliberal experiment which has shaped the bipartisan orientation of the country’s political landscape to this day (Kelsey, 1995). Income inequalities increased during the neoliberal reforms of the 1980s and 1990s (Carroll et al., 2011). In 2004, household incomes in the top 20% were five times higher than the bottom 20% (Ministry of Social Development, 2007). From 1982 to 2004, New Zealand ranked seventh out of twenty-five other OECD (Organisation for Economic Co-operation and Development) countries in terms of income inequality (Ministry of Social Development, 2007). Further, over this period, the nation witnessed the largest growth in inequalities among the OECD represented by a shift in New Zealand’s Gini coefficient from two points below the OECD average to three points above (Ministry of Social Development, 2007). It seems reasonable to suggest that this may have raised class awareness in New Zealand.

An increasing awareness of class however, contradicts theorists arguing that post-modernity has all but diminshed the impact of class position on individual social sentiments (Beck and Beck-Gernsheim, 2002; Kingston, 2000; Pakulksi, 2005; Pakulksi and Waters, 1996). This inconsistency is often used by others to construct an explanation of a ‘paradox of class’ (Bottero, 2004) or an ‘epistemological fallacy’ (Furlong and Cartmel, 2007), which involves the continuation of class in shaping life chances, while weakening subjectively as individualistic values proliferate. Discussions on the significance of class have recently intensified, with many pointing to empirical studies which question the notion that class has subjectively dwindled under neoliberalism (Caínzos and Voces, 2010; Evans and Tilley, 2012; Kelley and Evans, 1995; Hout, 2008). While many scholars have established the link between class and self-placements by comparing identifications using data from a number of countries (see Andersen and Curtis, 2012; Evans and Kelley, 2004; Oddsson, 2010), it is also important to acknowledge that variances in patterns of association between class position and subjective location are formed by the specific social context within which they occur (Edlund, 2003). Through a theoretically grounded and empirically tested analysis, I will illuminate how the class structure of New Zealand society continues to impinge on individual subjectivities.

Class identification

The seminal work of Richard Centers (1953, 1961) represents one of the earliest sociological enquiries into the link between ‘objective’ and subjective class (Ekehammar et al., 1987). Centers found that individuals in different occupations systematically placed themselves into different social classes, although these aspects were not necessarily perfectly correlated. Others (Goldthorpe et al., 1970; Jackman and Jackman, 1973; Marshall et al., 1988; Wright, 1997) have also found that most individuals understand social classes to exist, and most are able to self-place accordingly.

Class identification, the primary dependent variable of this study, is mainly concerned with an individual’s perception of his or her own position within a hierarchy stratified in class terms, whereas class consciousness represents an individual’s self-concept deriving from their knowledge of their membership in a social group, including knowledge of the values and emotions which are attached to the group. A more effective way to differentiate between the two closely related concepts is to regard class consciousness as containing a number of levels of which class identification could be regarded as the most fundamental (Carchedi, 1998; Lukács, 1968; McNamee and Vanneman, 1983;Vanneman, 1980).

Death of class?

Some commentators argue that class identities are dead. Critics in this vein contend that there are other more important sources of interests which influence identification. It is theorized that society has disappeared from the scene and thus feelings of doubt are no longer associated with structures of class, rather they are problems with the structures of individual psyches. Proponents of this idea adhere to the notion that there is a divergence between subjective features (political attitudes, self-placements, lifestyles, and leisure interests) and objective factors (societal position); in other words, a detachment of class from class position. Some scholars regard the weakness of class identities as indicative of a period of major social and cultural change that is decidedly different from the classed societies of the past (Giddens, 1973; Clark and Lipset, 1991; Pakulski and Waters, 1996; Kingston, 2000; Pakulski, 2005).

Arguably, the key proponent for the ‘death of class’ theory is Ulrich Beck. Beck and Beck-Gernsheim’s (2002) theory of ‘individualization’ asserts that individuals have been removed from status-based classes. Individuals are said to have been ‘disembedded’ from traditional communal modes through welfare-state policies and re-embedded into new modes in which the ability to create life paths and new identities is tantamount to individual ‘reflexivity’. The new ways of living under postmodernity have apparently revealed ‘dynamic possibilities for a reorganization of social relations, which cannot be adequately comprehended by following either Marx or Weber’ (Beck and Beck-Gernsheim, 2002: 36).

Analyses testing the argument put forth by those who believe emerging postmodern identities undermine the impact of class on individual subjectivities, have revealed that competing identities often do not rival those of social class (Andersen and Curtis, 2012; Atkinson, 2007a, 2007b, 2007c, 2010a, 2010b; Evans et al., 1992; Goldthorpe and McKnight, 2006; Hout, 2008; Evans and Kelley, 2004; Marshall et al., 1988). In terms of political behaviour, the continuing relevance of class as a source of differentiation has been demonstrated in 22 European countries (Caínzos and Voces, 2010). There are certainly class practices which structure society objectively and subjectively, but there is often a denial or an unwillingness to situate oneself within this structure. Often the inability to identify with the ‘correct’ class is associated with a psychological factor known as reference group theory (Merton, 1957; Shibutani, 1955; Siegel and Siegel, 1957). Reference group theory is concerned with the relationship between structural integration and structural perception, whereby ‘an actor’s structure perception is conditioned by his or her structural neighbourhood’ (Levy, 1991: 65). Reference group processes are associated with the ‘middling’ of subjective social positions (Evans et al., 1992; Evans and Kelley, 2004; Kelley and Evans, 1995). It is, in part, the ‘middling’ of subjectivities which leads reflexivity theorists to regard class as an insignificant feature of identity and society.

Research question

What is the nature of the relationship between an individual’s class position in New Zealand society and the subjective perceptions of these positions? In light of the current debates within the literature surrounding the death of class and answer to such a question has broader methodological implications for the future of class analysis and can also be used to assess the relevance of Weber’s conception of class situation. For many scholars, there is an increasing gap between class conditions in terms of structure and the subjective perceptions of these conditions. Bottero (2004: 988), for example, suggests that a ‘curious symmetry’ has emerged whereby post-modern writers have ‘abandoned economic relations in their attempt to explain social identity’, and traditional class analysts ‘have jettisoned social identity as a key component of class analysis’. This study represents an attempt to reintroduce symmetry between the fields of thought.

Economic relations remain pertinent in shaping self-identification, which I demonstrate by incorporating individual identification into an analysis of social class. It is important to determine the mechanisms which orient subjectivities alongside social structural processes, and in turn how these structural processes affect and impact on individual subjectivity.

Data and methods

The data employed consists of two national representative samples derived from the International Social Survey Programme (ISSP). The first, in 1999, was gathered at a time when citizens were concerned with the social impacts of neoliberal reforms (see Humpage, 2011).The second, in 2009, was collected shortly after the global economic crisis struck near the end of 2008. Taken together, the data provides a unique snapshot of New Zealand during a period when its citizens were exposed to, and later experienced the ramifications of neoliberal rationality. Including this time dimension also allows for an assessment of the changing predictive power and salience of class over the decade.

The ISSP is an annually run social survey conducted in over 30 countries worldwide. Between June and August 1999, a nationwide postal survey was conducted on 2100 randomly selected individuals over the age of 18 from the New Zealand electoral roll (Gendall et al., 2000). The 1999 survey produced produced 1118 valid responses, which equates to a response rate of 61%. The final sample size of the actual data set used is 1108. The 2009 iteration was collected between July and November of the same year using the same method of collection. The survey generated 935 valid responses, which is an effective response rate of 44%. Both data sets were compared to the closest year’s census figures and are generally representative. However in terms of age, young people are under-represented as the survey only includes those over the age of 18. In addition, there is a slightly higher proportion of women and a slightly lower proportion of men,

Dependent variable: subjective class

Subjective class was operationalized in both data sets using a common question which queried respondents on their location according to six class categories: (1) ‘Lower class’; (2) ‘Working class’; (3) ‘Lower middle class’; (4) ‘Middle class’; (5) ‘Upper middle class’; and (6) ‘Upper class’. Self-placement questions of this sort have been used extensively by others when measuring class identification (Davis and Robinson, 1988; Evans and Mills, 1998; Jackman and Jackman, 1973; Robinson and Kelley, 1979). By providing six possible choices and including a hierarchical dimension, this question takes on an ordinal scale form as opposed to a nominal categorical ranking.¹ As the number of ‘Upper class’ responses is too small for a detailed analysis, the associated values were combined with the ‘Upper middle class’ category

Primary individual level predictors: class, education and income

Social class, education and income are regarded as the primary ‘objective’ characteristics in this empirical investigation to predict subjective class identification. The decision to measure the effects of class, education and income separately was taken as it is expected that each may impact independently on self-placement (Chan and Goldthorpe, 2007). Factitious measures of socioeconomic position combining education, occupation and income into a composite measure of class worry many analysts as their various interconnections are hidden during multivariate analysis. A number of other scholars researching class support this position. Kelley and Evans (1995) for instance, contend that education and income merely reflect a distinctive element of class. Likewise, Evans and Mills (1998: 87) argue for the separation of class, education and income, as these ‘may be of interest in their own right, as explanatory dependent variables’. Conley (2008: 371) sums up the paradox of class and status rather succinctly:

the moment we are measuring some aspect of [social class] adequately – say, income – and specifying a causal pathway of some sort, then we have taken our finger off what class really is … it is the silent force between the cracks of wealth, income, occupation, and education that constitute the mortar of the class system.

Therefore, treating class and aspects of socioeconomic position as separate explanatory variables, and measuring the impact each has on self-placement could provide insights as to which exerts the greatest effect.

Class was ‘operationalized’ using a variation of the Goldthorpe (EGP) class schema known as the European Socioeconomic Classification (ESeC) (Rose and Harrison, 2010). The ESeC, which is inspired by a Weberian interpretation of class, is regarded as a newer version of Goldthorpe’s schema, carrying the ‘essential genes of a highly regarded, tried-and-trusted social class measure’ (Harrison and Rose, 2006: 291). As those in similar class positions possess certain resources by virtue of the positions they occupy, it is understood that they will share comparable possibilities and restrictions in relation to ‘life-chances’. Arguably, Goldthorpe’s schema is now the most influential categorical measure of social class and is often regarded as the most valuable way of measuring, empirically, class-based inequalities (Evans, 1996; Le Roux et al., 2008; Marks, 1999, 2005; Marshall et al., 1988; Svallfors, 2005). The schema may not be perfect but it is widely used and contains both construct and criterion related validity (Bergman and Joye, 2005; Evans, 1992; Evans and Mills, 1998, 2000), and can be augmented to counteract data deficiencies (Caínzos and Voces, 2010; Güveli et al., 2007). It provides a useful list of class situations when one wishes to assess the causal significance of these in relation to self-placement.

Education and income ‘have long been recognized as important determinants of class identification’ (Robinson and Kelley, 1979: 52). Researchers often incorporate number of years of schooling as a measurement of individual educational level (Ganzeboom et al., 1992). However, it makes more sense to incorporate highest credential obtained, as a student could attend an institution for a number of years and still not acquire certification (see Castillo, 2012). Table 1 provides the percentage distributions for education. Most had obtained below a higher secondary credential in 1999; in 2009 the majority of the sample had obtained an above higher secondary credential. The median level in 1999 was ‘Higher secondary’; in 2009 the level had increased to ‘Above higher secondary’. The percentage of those who had obtained an above higher secondary or university credential has significantly increased from 1999 to 2009.

Table 1.

ISSP 1999/2009 educational distributions.

Level of education	1999	2009
Below higher secondary	41.7%	25.1%
Higher secondary	27.4%	19.0%
Above higher secondary	16.3%	30.2%
University	14.7%	25.7%
	100.0	100.0
Missing cases	19	9
Total	1089	926

As we know from Marx (1893: 862–3, cited in Robinson and Kelley, 1979: 42), social classes should not be equated with income groupings as classes are not defined by the ‘sources of revenue’ or the ‘size of purses’. Aside from this, however, it is expected that class and income, to a degree, should be correlated. Respondents in both years were asked to select from a number of categories which best described their personal yearly income from all sources before tax. The income distributions of both datasets are provided in Table 2. Most earned ‘$10,000 or less per year’ in 1999; in 2009 the majority were earning between ‘$50,001 and $70,000’. In 1999 the median income bracket was ‘$25,001– $30,000’; in 2009 this had shifted upwards to ‘$30,001–$40,000’.

Table 2.

ISSP 1999/2009 income distributions.

Income categories	1999	2009
$10,000 NZD or less per year	16.7%	10.5%
$10,001– $15,000 NZD	12.3%	7.0%
$15,001–$20,000 NZD	9.8%	8.3%
$20,001–$25,000 NZD	9.4%	7.9%
$25,001–$30,000 NZD	9.5%	8.7%
$30,001–$40,000 NZD	15.1%	12.7%
$40,001–$50,000 NZD	11.2%	12.8%
$50,001–$70,000 NZD	9.4%	14.4%
$70,001–$100,000 NZD (or more per year in 1999)	6.6%	10.7%
$100,001 NZD or more per year	–	6.8%
	100.0	100.0
Missing cases	55	40
Total	1053	895

Findings

Table 3 provides a tabular illustration of the subjective class responses in terms of percentages for both periods. The relatively low missing values, particularly in the latter period, are illustrative of the solid position of class within identity formation. Arguably, then, individual perceptions of class are sufficiently solid to allow it to be seen as one of many possible sources of identity. In 1999, the median value was the ‘Middle class’ category. In 2009, most respondents also perceived themselves to belong to the ‘Middle class’; although this number has increased by approximately 5% over the decade.

Table 3.

ISSP 1999/2009 subjective class distributions.

Subjective class responses	1999	2009
Lower class	5.1%	3.4%
Working class	17.6%	15.9%
Lower middle class	15.2%	15.2%
Middle class	46.5%	51.5%
Upper middle class	15.6%	14.1%
	100.0	100.0
Missing cases	122	11
Total	989	801

To improve the readability of the results and in order to have enough respondents in each class for a detailed analysis, the original nine-class schema was collapsed down to a five-class model (see Evans, 1992). The distributions of the collapsed version can be seen in Figures 1 and 2 below. The allocations here are used for the analyses to follow. Figures 1 and 2 illustrate graphically the subjective class placements of individuals in different classes for 1999 and 2009. For the most part, ‘middle class’ placements dominated the responses to subjective class in both periods. That said, class identifications are also structured by class. Both figures rank the classes from highest percentage ‘working class’ response to lowest. The figures depict that in both years,besides ‘Skilled, semi-, and non-skilled workers’, strong majorities of all other classes identified with the ‘middle class’. In addition, ‘middle class’ placements are the highest with the ‘Salariat’ in both 1999 and 2009, and are at the lowest with the ‘Skilled, semi-, and non-skilled workers’. As well, in 2009 the ranking of ‘middle class’ responses closely follows (besides the positions of Class 2 and 3) the order of the class schema if it is treated in hierarchical terms. This demonstrates that while most identify with the ‘middle class’, the responses are structured in terms of class. It is also evident that in both years ‘working class’ identifications correlate with class; the percentages increase as we shift from Class 1 to Class 5; whereas the reverse is true for ‘upper middle class’ identifications in both years.

Figure 1.

Bar chart of subjective class by objective class 1999 (n = 893).

Figure 2.

Bar chart of subjective class by objective class 2009 (n = 794).

We have established that there appears to be a relationship between the objective predictor variables and the subjective outcome variable; yet the specific nature of these linkages and the predictor variable(s) that exerts the strongest influence on self-placement remains to be seen. In order to compare the relationships between objective aspects of stratification and subjective class, we need an index of the degree of strength of the relationship between the variables that is comparable from one pair to another. Moreover, additional questions are raised: should the associations be considered weak or strong? How great an increase is found in subjective class for a given increase in education or income, or when we shift from one class group to another? Answers to these questions can partially be obtained through correlation and regression analyses.

Correlations between self-placement and objective indicators such as occupation, income, and education have commonly ranged from approximately .20 to .50 (Ostrove et al., 2000). Coefficients between occupation, education and class identification have typically ranged from approximately .3 to .5; between income and class identification they have typically fallen in between .3 and .4 (Goyder, 1975). Ekehammar et al. (1987) found that the magnitude of the correlation among indicators of socioeconomic position and the subjective aspects of class was approximately .3, with positive and often statistically significant relationships.

Correlation coefficients used independently do not provide an indication of the direction of causality. Namely, although we can state that as we move from Class 1 to Class 5, subjective class and status decrease, or that self-placements increase alongside income and education, we cannot state simply through the use of a correlation coefficient that these objective indicators cause subjective class and status to increase. This is because either a third measure or an unmeasured variable may be affecting the results. Moderate to large coefficient values do not ‘prove’ causal assumptions. They do however, make them tentatively more plausible. That said, theoretically, our understanding is that objective elements such as class, income and education, are associated with subjective class. In this way, causation entails ‘predictability in accordance with theory’ (Goldthorpe, 2001: 3).

Tables 4 and 5 provide a correlation matrix for the variables of interest in 1999 and 2009. The class coefficients have remained relatively constant during the 10-year period as income and education have fluctuated. In the earlier period, education and income were stronger predictors of subjective class than the ‘objective’ class measurement. A decade later, education and income have declined in strength as a predictor of class identification. The effect of class on the subjective element is not consistently the strongest when compared to the components of socioeconomic position, yet it has remained relatively stable over the 10-year period, as income and education have declined. The increase in the strength of the correlation between class, income and education indicates that the latter two may be of interest in their own right as explanatory dependent variables.

Table 4.

Correlation matrix – 1999.

	1999
	Income	Education	Subjective class	Class
Income	1	.24^*	.38^*	.38^*
Income	1053	1039	946	947
Education		1	.32^*	.32^*
Education		1089	974	972
Subjective class			1	.34^*
Subjective class			986	893
Class				1
Class				989

p < .001 (2-tailed).

Table 5.

Correlation matrix – 2009.

	2009
	Income	Education	Subjective class	Class
Income	1	.29^*	.26^*	.43^*
	895	888	888	776
Education		1	.26^*	.45^*
		926	918	796
Subjective class			1	.36^*
			924	794
Class				1
				801

p < .001 (2-tailed).

Table 6 regresses class identification on class (treated as both a nominal and ordinal measurement), income, and education. Gender and age are included as control variables. The impact of the model was greater in 1999 than in 2009. That is to say, in 1999 the model explained 28% of the variance in responses to subjective class; in 2009 the model explained 17% of the variance. This can be explained because the impacts of income and education declined in 2009 as evidenced by the correlation coefficients in Tables 4 and 5. The effects of class, income and education on subjective class are all in the predicted direction and all are statistically significant (p < .01).² The positive and significant regression coefficients of income and education in both periods indicate that the higher one’s annual income or education, the more likely one is to identify with a higher class.

Table 6.

Self-placement regressed onto class, income, education, gender, and age (unstandardized beta coefficients including associated standard errors).

	1999		2009
	B	SE B	B	SE B
Constant	2.01	.20	2.85	.25
Class	.14***	.02	.18***	.03
Income	.14***	.02	.04**	.01
Education	.24***	.03	.15***	.04
Gender (male=0, female=1)	.15*	.07	−.03	.07
Age	.01***	.00	.01***	.00
R ²	.28	–	.18	–
Adj. R²	.28	–	.17	–
N	824	–	767	–
Constant	1.78	.19	2.64	.20
Intermediate employees (Class 2)	−.05	.11	−.18	.09
Small employers and self-employed (Class 3)	−.12	.10	−.16	.12
Lower grade white collar workers (Class 4)	−.06	.14	–.64***	.13
Skilled, semi- and non-skilled workers (Class 5)	−.61***	.10	−.69***	.12
Income	.14***	.02	.04*	.02
Education	.25***	.03	.15***	.04
Gender	.14	.07	−.001	.07
Age	.01***	.002	.01***	.002
R ²	.29	–	.18	–
Adj. R²	.29	–	.18	–
N	824	–	767	–

***

Significance at p < .001; **Significance at p < .01; *Significance at p < .05.

The class variable is usually treated nominally (see Erikson and Goldthorpe, 1992; Wright, 1997), therefore the five-class schema was ‘dummy’ coded so that it is more suitable for regression analysis. The ‘Salariat’ was chosen as the referent category or baseline group as this group represents the class with the majority of people. As well, it is useful to compare the self-placements of the most ‘advantaged’ class to the most ‘disadvantaged’. The negative values indicate that subjective placements decrease in all other classes when compared to the ‘Salariat’.

In both periods, the unstandardized coefficients for Classes 2 through 5 (listed in the lower half of the table) are negative which indicates that as we move from the most ‘advantaged’ class to the most ‘disadvantaged’ class, self-placements decrease. Taking the subjective class coefficient of ‘Skilled, semi- and non-skilled workers’ in 2009 as an example; the value of -.69 indicates that, net of the other variables, when compared to the ‘Salariat’, this class self-identified slightly less than one category below the ‘Salariat’ when answering the self-placement question.³

In 2009, the coefficients of the specific classes remain negative, indicating that as we move from Class 1 to 5, subjective class decreases. There is little difference between the subjective class coefficients for Classes 2, 3 and 4 in 1999, which indicates that these groups have lower subjective class identifications than the ‘Salariat’ class and they tend to be similar to each other. In 2009 however, there is a clearer divide between Classes 1 to 3 and Classes 4 to 5. In 2009, the class coefficients for Class 2 and 3 are now similar, whereas the values for Class 4 are now closer to Class 5. In other words, in 2009, all classes held lower self-placements than the ‘Salariat’; yet when compared to this group, those in Class 4 and 5 had significantly lower class identifications than those in Class 2 and 3. If we turn our attention solely to ‘Skilled, semi- and non-skilled workers’ in both periods, one notices that class identifications were lower in this class than in the other four. The relationship between class position and self-placements is clearest between the ‘Salariat’ and the ‘Skilled, semi- and non-skilled workers’. The adjusted R² values, suggest that the explanatory power of class is modest. Class is not the only structuring influence on class identification, but it has remained as a relatively modest yet resolute predictor.

It has been demonstrated that class certainly does not explain all of the variation in subjective self-placement; rather there is a complex relationship between class, income and education. Because of the reduction in the strength of the correlation coefficients when aspects of socioeconomic position are ‘controlled for’, it is suspected that there is a direct and indirect causal relationship between class situation and class identification, and that this relationship persists in both periods. The initial relationship between class position and subjective class is partially due to the class differences in income and educational levels. If all classes earned the same annual income and had obtained the same level of education, the effect of class on self-placements would be less but would still remain.

When a time dimension is included, alongside other structural cleavages, class situation remains as a significant source of self-placement. As a predictor of class identification, the measurement of objective class has retained a moderately significant correlation as the connections between this form of self-placement and other aspects of stratification have declined. These findings demonstrate that class is relatively important and has some influence on structuring identities. That said, the relationship between class and self-placement is further complicated by reference group forces.

There is a definite middling tendency, yet in both periods the most ‘disadvantaged’ classes all had significantly lower self-placements than did the most ‘advantaged’ classes. In this regard, it is notable that the influence of class on subjectivities is strongest in those most greatly affected by class cleavages. Likewise, in both periods, those with the lowest levels of education and income frequently held lower self-placements when compared to those of higher levels. These results demonstrate that the objective structuring of subjective social locations is more evident at the extremes of class and status – the relationship between objective social position and self-placement is crystallized at the limits of class, education and income.

Discussion and conclusion

The present study has furthered an understanding of the nature of the relationship between class situation and class identification in New Zealand. The findings appear to add credence to Weber’s theory that class bearings may overshadow status placements during periods of economic upheavals. What is gleaned from the correlation between class identification, status groups and class situation is that occupationally derived measures of class, rather than measures derived from income and education, continue to have purchase. This inference, combined with the substantial decrease in the non-responses over the 10-year period, leads me to deduce that class has certainly not diminished as a source of identity as others have claimed.

A less than perfect fit between ‘objective’ and subjective class does not imply that the impact of the former has diminished. The issue is not explanatory reductionism versus individual subjective autonomy, but rather the impact of different causal factors and how they are interrelated. Class identification may not always occur on the basis of similar objective situations, but occupying these situations creates the potential for similar identifications. As social identities are in part constructed in class terms, the claim that there are no grounds for asserting that ‘social class is entirely opaque to the majority of individuals and therefore unavailable as a source of social identity’ is strongly supported (Marshall et al., 1988: 145). In a New Zealand context, social class position continues to a have real, if variable, impact on individual subjectivities.

However, an issue remains as to whether responses to subjective class questions make substantive sense or whether they represent little more than ‘random acquiescence to the survey situation’ (Hout, 2008: 31). The latter would suggest little class awareness on the part of the individual. In other words, this analysis has not explained how far indicators of class perceptions capture what social class fully means for respondents. The difficulty with measuring subjective class quantitatively is that often responses are directed into channels preconceived by survey designers or investigators. In other words, ‘adopting a mode of questioning which obliges a response in terms of the researcher’s system of analysis whether or not this had previously been shared by the respondent’ (Child et al., 1980: 372).

Obtaining subjective perceptions of class through surveys raises an issue relating to whether class is actually employed by respondents outside in their daily lives, or whether they are merely induced to conceive of themselves in class terms. Acknowledging that this may be the case, it still remains pertinent to first establish whether or not these inducements are structured in terms of class on a broader societal level, and then to ascertain the nuances of the impact social class has and continues to have in the everyday lives of individuals, including how they characterize themselves in relation to others. Further qualitative studies could supplement quantitatively derived findings with respect to how individuals construct class identity. Quantitative indicators of class perceptions do demonstrate the salience of class, yet they do not capture what class fully means to respondents. They do however produce a sound nomothetic basis for future qualitative research which addresses class as a subjective phenomenon.

The global economic crisis may have raised class awareness as citizens upset with austerity measures have increasingly become more critical of their society. A potential manifestation of this is that individual perceptions of class and class divisions are increasingly unobstructed. As Atkinson (2010b: 427) suggests, ‘the substance of class may have altered over time but the relations structuring the objective and subjective field of possibles remain as obdurate as ever’. In times of varied and fluid social awareness, the homogenizing conditions of class have remained as salient sources of individual identity.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Author biography

Edward Haddon is pursuing his PhD in the Department of Sociology at the University of British Columbia. His current work explores the subjective implications of social class using mixed-method approaches, youth subjectivities, and classical and contemporary theory, specifically related to social stratification.

References

Andersen

Curtis

(2012) ‘The Polarizing Effect of Economic Inequality on Class Identification: Evidence from 44 Countries’, Research in Social Stratification and Mobility 30(1): 129–41.

Atkinson

(2007a) ‘Anthony Giddens as Adversary of Class Analysis’, Sociology 41(3): 533–49.

Atkinson

(2007b) ‘Beck, Individualization and the Death of Class: A Critique’, British Journal of Sociology 58(3): 349–66.

Atkinson

(2007c) ‘Beyond False Oppositions: A Reply to Beck’, British Journal of Sociology 58(4): 707–15.

Atkinson

(2010a) Class, Individualization and Late Modernity: In Search of the Reflexive Worker. New York: Palgrave Macmillan.

Atkinson

(2010b) ‘The Myth of the Reflexive Worker: Class and Work Histories in Neo-liberal Times’, Work, Employment and Society 24(3): 413–29.

Beck

Beck-Gernsheim

(2002) Individualization: Institutionalized Individualism and its Social and Political Consequences. London: Sage.

Bergman

M.M.

Joye

(2005) ‘Comparing Social Stratification Schemata: CAMSIS, CSP-CH, Goldthorpe, ISCO-88, Treiman, and Wright’, Cambridge Studies in Social Research 10. URL (consulted March 2012): www.sociology.cam.ac.uk/srg/workingpapers/CS10StratificationSchemata2.pdf

Bottero

(2004) ‘Class Identities and the Identity of Class’, Sociology 38(5): 985–1003.

10.

Caínzos

Voces

(2010) ‘Class Inequalities in Political Participation and the “Death of Class” Debate, International Sociology 25(3): 383–418.

11.

Carchedi

(1998) ‘Classes and Class Analysis’, pp. 105–25 in Wright

E.O.

(ed.) The Debate on Classes. London: Verso.

12.

Carroll

Casswell

Huakau

Howden-Chapman

Perry

(2011) ‘The Widening Gap: Perceptions of Poverty and Income Inequalities and Implications for Health and Social Outcomes’, Social Policy Journal of New Zealand 37: 111–22.

13.

Castillo

J.C.

(2012) ‘Is Inequality Becoming Just? Changes in Public Opinion about Economic Distribution in Chile’, Bulletin of Latin American Research 31(1): 1–18.

14.

Centers

(1953) ‘Social Class, Occupation, and Imputed Belief’, American Journal of Sociology 58(6): 543–55.

15.

Centers

(1961) The Psychology of Social Classes: A Study of Class Consciousness. New York: Russell and Russell.

16.

Chan

T.W.

Goldthorpe

J.H.

(2007) ‘Class and Status: The Conceptual Distinction and its Empirical Relevance’, American Sociological Review 72(4): 512–32.

17.

Child

Pearce

King

(1980) ‘Class Perceptions and Social Identification of Industrial Supervisors’, Sociology 14(3): 363–99.

18.

Clark

T.N.

Lipset

S.M.

(1991) ‘Are Social Classes Dying?’, International Sociology 6(4): 397–410.

19.

Conley

(2008) ‘Reading Class Between the Lines (of this volume): A Reflection on Why We Should Stick to Folk Concepts of Social Class’, pp. 366–73 in Lareau

Conley

(eds) Social Class: How Does It Work? New York: Russell Sage Foundation.

20.

Davis

N.J.

Robinson

R.V.

(1988) ‘Class Identification of Men and Women in the 1970s and 1980s’, American Sociological Review 53(1): 103–12.

21.

Edlund

(2003) ‘The Influence of the Class Situations of Husbands and Wives on Class Identity, Party Preference and Attitudes Towards Redistribution: Sweden, Germany and the United States’, Acta Sociologica 46(3): 195–214.

22.

Ekehammar

Sidanius

Nilsson

(1987) ‘Social Status: Construct and External Validity’, Journal of Social Psychology 127(5): 473–81.

23.

Erikson

Goldthorpe

J.H.

(1992) The Constant Flux: A Study of Class Mobility in Industrial Societies. New York: Clarendon Press.

24.

Evans

(1992) ‘Testing the Validity of the Goldthorpe Class Schema’, European Sociological Review 8(3): 211–32.

25.

Evans

(1996) ‘Putting Men and Women into Classes: An Assessment of the Cross-sex Validity of the Goldthorpe Class Schema’, Sociology 30(2): 209–34.

26.

Evans

Mills

(1998) ‘Identifying Class Structure: A Latent Class Analysis of the Criterion-related and Construct Validity of the Goldthorpe Class Schema’, European Sociological Review 14(1): 87–106.

27.

Evans

Mills

(2000) ‘In Search of the Wage-labour/service Contract: New Evidence on the Validity of the Goldthorpe Class Schema’, British Journal of Sociology 51(4): 641–61.

28.

Evans

Tilley

(2012) ‘How Parties Shape Class Politics: Explaining the Decline of the Class Basis of Party Support’, British Journal of Political Science 42(1): 137–61.

29.

Evans

M.D.R.

Kelley

(2004) ‘Subjective Social Location: Data from 21 Nations’, International Journal of Public Opinion Research 16(1): 3–38.

30.

Evans

M.D.R.

Kelley

Kolosi

(1992) ‘Images of Class: Public Perceptions in Hungary and Australia’, American Sociological Review 57(4): 461–82.

31.

Furlong

Cartmel

(2007) Young People and Social Change: New Perspectives. Maidenhead: Open University Press.

32.

Ganzeboom

H.B.G.

De Graaf

P.M.

Treiman

D.J.

(1992) ‘A Standard International Socio-economic Index of Occupational Status’, Social Science Research 21(1): 1–56.

33.

Gendall

Robbie

Patchett

Bright

(2000) Social Equality in New Zealand: The International Social Survey Programme. Palmerston North: Massey University.

34.

Giddens

(1973) The Class Structure of Advanced Societies. London: Hutchinson.

35.

Goldthorpe

J.H.

(2001) ‘Causation, Statistics, and Sociology’, European Sociological Review 17(1): 1–20.

36.

Goldthorpe

J.H.

McKnight

(2006) ‘The Economic Basis of Social Class’, pp. 109–36 in Morgan

S.L.

Grusky

D.B.

Fields

G.S.

(eds) Mobility and Inequality: Frontiers of Research in Sociology and Economics. Stanford, CA: Stanford University Press.

37.

Goldthorpe

J.H.

Lockwood

Bechhofer

Platt

(1970) The Affluent Worker: Industrial Attitudes and Behaviour. Cambridge: Cambridge University Press.

38.

Goyder

J.C.

(1975) ‘A Note on the Declining Relation between Subjective and Objective Class Measures’, British Journal of Sociology 26(1): 102–9.

39.

Güveli

Need

De Graaf

N.D.

(2007) ‘The Rise of “New” Social Classes within the Service Class in the Netherlands’, Acta Sociologica 50(2): 129–46.

40.

Harrison

Rose

(2006) The European Socio-economic Classification (ESeC) User Guide. Institute for Social and Economic Research, University of Essex. URL (consulted April 2012): www.iser.essex.ac.uk/files/esec/guide/docs/UserGuide.pdf

41.

Hout

(2008) ‘How Class Works: Objective and Subjective Aspects of Class Since the 1970s’, pp. 25–64 in Lareau

Conley

(eds) Social Class: How Does it Work? New York: Russell Sage Foundation.

42.

Humpage

(2011) ‘Neo-liberal Reform and Attitudes towards Social Citizenship: A Review of New Zealand Public Opinion Data 1987–2005’, Social Policy Journal of New Zealand 37: 83–96.

43.

Jackman

M.R.

Jackman

R.W.

(1973) ‘An Interpretation of the Relation between Objective and Subjective Social Status’, American Sociological Review 38(5): 569–82.

44.

Kelley

Evans

M.D.R.

(1995) ‘Class and Class Conflict in Six Western Nations’, American Sociological Review 60(2): 157–78.

45.

Kelsey

(1995) The New Zealand Experiment: A World Model for Structural Adjustment? Auckland: Auckland University Press.

46.

Kingston

P.W.

(2000) The Classless Society. Stanford, CA: Stanford University Press.

47.

Le Roux

Rouanet

Savage

Warde

(2008) ‘Class and Cultural Division in the UK’, Sociology 42(6): 1049–71.

48.

Levy

(1991) ‘Structure-blindness: A Non-ideological Component of False Consciousness’, International Journal of Sociology and Social Policy 11(6–8): 61.

49.

Lukács

(1968) History and Class Consciousness: Studies in Marxist Dialectics. Cambridge, MA: MIT Press.

50.

Marks

G.N.

(1999) The Measurement of Socioeconomic Status and Social Class in the LSAY Project. Australian Council for Educational Research (ACER). URL (consulted April 2012): www.acer.edu.au/documents/LSAY_techrep14.pdf

51.

Marks

G.N.

(2005) ‘Cross-national Differences and Accounting for Social Class Inequalities in Education’, International Sociology 20(4): 483–505.

52.

Marshall

Newby

Rose

Vogler

(1988) Social Class in Modern Britain. London: Hutchinson.

53.

Ministry of Social Development (2007) Household Incomes in New Zealand:Trends in Indicators of Inequality and Hardship 1982–2004. URL (consulted August 2012): http://www.msd.govt.nz/about-msd-and-our-work/publications-resources/monitoring/household-incomes/household-incomes-1982–2004.html

54.

McNamee

S.J.

Vanneman

(1983) ‘The Perception of Class: Social and Technical Relations of Production’, Work and Occupations 10(4): 437–69.

55.

Merton

R.K.

(1957) Social Theory and Social Structure. Illinois: Free Press of Glencoe.

56.

Oddsson

G.A.

(2010) ‘Class Awareness in Iceland’, International Journal of Sociology and Social Policy 30(5/6): 292–312.

57.

Ostrove

J.M.

Adler

N.E.

Kuppermann

Washington

A.E.

(2000) ‘Objective and Subjective Assessments of Socioeconomic Status and their Relationship to Self-rated Health in an Ethnically Diverse Sample of Pregnant Women’, Health Psychology 19(6): 613–18.

58.

Pakulski

(2005) ‘Foundations of a Post-class Analysis’, pp. 152–279 in Wright

E.O.

(ed.) Approaches to Class Analysis. Cambridge: Cambridge University Press.

59.

Pakulski

Waters

(1996) ‘The Reshaping and Dissolution of Social Class in Advanced Society’, Theory and Society 25(5): 667–91.

60.

Robinson

R.V.

Kelley

(1979) ‘Class as Conceived by Marx and Dahrendorf: Effects on Income Inequality and Politics in the United States and Great Britain’, American Sociological Review 44(1): 38–58.

61.

Rose

Harrison

(eds) (2010) Social Class in Europe: An Introduction to the European Socio-Economic Classification. Abingdon: Routledge.

62.

Savage

Devine

Cunningham

Taylor

Hjellbrekke

. (2013) ‘A New Model of Social Class? Findings from the BBC’s Great British Class Survey Experiment’, Sociology 47(2): 219–50.

63.

Shibutani

(1955) ‘Reference Groups as Perspectives’, American Journal of Sociology 60(6): 562–9.

64.

Siegel

A.E.

Siegel

(1957) ‘Reference Groups, Membership Groups, and Attitude Change’, Journal of Abnormal and Social Psychology 55(3): 360–4.

65.

Svallfors

(2005) ‘Class and Conformism: A Comparison of Four Western Nations’, European Societies 7(2): 255–86.

66.

Vanneman

R.D.

(1980) ‘U.S. and British Perceptions of Class’, American Journal of Sociology 85(4), 769–90.

67.

Weber

(2000 [1914]) ‘Class, Status, Party’, pp. 114–24 in Grusky

D.B.

(ed.) Social Stratification: Class, Race, and Gender in Sociological Perspective. Boulder, CO: Westview Press.

68.

Weeden

K.A.

Grusky

D.B.

(2012) ‘The Three Worlds of Inequality’, American Journal of Sociology 117(6): 1723–85.

69.

Wright

E.O.

(1997) Classes. London: Verso.