70
Social Mobility and Education in Britain
children - against it have to be seen as creating a large potential for resistance to any attempts at widening opportunity, in the sense of equalising mobility chances, whether through educational policy or otherwise.
It is in this regard instructive that in the one instance we have found of relative rates in fact becoming more equal - that is, among women part-timers - it would appear that this trend results from growing numbers of women from more advantaged class backgrounds in effect opting to become downwardly mobile intergenerationally by declining opportunities in the labour market - if not in the marriage market -that their class origins and educational attainment would probably have made available to them. But, in general, it has to be accepted that policies aimed at reducing inequalities in relative mobility chances will not be costless to the members of families who hold more advantaged class positions; and that, rather than showing a merely passive acceptance, these families can be expected to apply their superior resources, economic and otherwise, in all available ways in order to counter or circumvent such policies.
The Pattern of Social Fluidity within the Class Structure: Hierarchy, Inheritance and Status Effects
In the previous chapter we have shown that over a period extending back at least to the middle of the last century there has been little change in relative rates of class mobility in British society. Or, in other words, little change has occurred in the level of social fluidity within the class structure as expressed by the strength of the association between the class positions of children and their parents when considered net of all structural change. The only exception arises in the rather special case of relative rates becoming more equal among women who at some point have worked part-time. In the present chapter we move on to consider a number of questions that arise. First, given the essentially stable level of social fluidity, so far as the large majority of the active population is concerned, what is the pattern of this fluidity within the class structure and how is this pattern created? Second, is this pattern itself stable over time? Third, is this pattern and its degree of stability the same for men and for those women who when in employment have only worked full-time? And, finally, and most consequentially, what are the implications of the patterning of social fluidity for the likelihood of different mobility transitions being made and in turn for issues of equality of opportunity?
When considering possible trends in the level of social fluidity, we observed that because relative mobility rates are captured by a very large number of different odds ratios, it is necessary to proceed by formulating statistical models that make statements about all odds ratios of interest and by then seeing how well these models can
1 We do not seek to address a comparable set of questions in the case of women who have worked part-time since we know that in their case the level of fluidity has in fact increased, so that in certain respects at least the pattern of fluidity cannot have remained stable - and also must be, or have become, different from that applying in the case of men and of women who have only worked full-time. To investigate in further detail the changes that have occurred would require more data on part-timers than we have presently available.
72
Social Mobility and Education in Britain
reproduce the empirical data. The same applies in treating the pattern of social fluidity.
In the light of theoretical arguments that have successfully guided previous research,2 we envisage a model according to which the pattern of social fluidity within the class structure is created by three different kinds of effect: those of class hierarchy, class inheritance and status affinity. We now explain these effects in turn.3
Hierarchy Effects. Hierarchy effects are ones that limit mobility between classes as a result of differences in the advantages or disadvantages that are associated with them as classes of origin - in terms of family economic, social and cultural resources; and of differences in the barriers that exist to their attainment as classes of destination - in terms of required skills, qualifications or capital. We noted in Chapter 1 that the seven NS-SEC classes we use in our analyses of mobility cannot be regarded as entirely hierarchically ordered but that, as was indicated in Table 1.1, four lines of hierarchical division can be drawn, with Classes 3, 4 and 5 being for this purpose placed at the same level (these were the divisions we previously used in Chapter 2 in order to define upward and downward as distinct from 'horizontal' mobility). Correspondingly, in our model we include four class hierarchy effects, labelled as HI1, HI2, HI3 and H14, that relate, as is shown in Table 4.1, to cells of the mobility table that involve the crossing of one, two, three or all four of the hierarchical divisions. These effects, as can be seen, are intended to operate cumulatively with the hierarchical range of the mobility transitions that are involved.
Inheritance Effects. Inheritance effects are ones that promote immobility and thus, like hierarchy effects, also limit mobility. This occurs, on the one hand, as a result of the distinctive motivations and opportunities that individuals may have for remaining in the same class as that in which they originated, as, say, through family occupational traditions or the direct intergenerational transmission of family businesses or capital; and, on the other hand, as a result of distinctive constraints that may exist on individuals' mobility away from their class of origin, as, say, through restricted employment possibilities in local labour markets. Inheritance effects thus apply only in cells on the main diagonal of the mobility table - that is, in those cells indicating immobility - and in our model
2 See in particular Erikson and Goldthorpe (1992: chs. 4 and 5).
3 For full technical details of the model - known as a topological model - that we go on to describe, see Bukodi, Goldthorpe and Kuha (2017).
.3 -a
a
s
ts K
I1
1
a
U
c So
o
"8
•J.
a 0
Ü
2
X
a
X
X
-
u
X
a a
s 2
□
x
2 a
X X
2   E   2   E x
E   E E
□
X
x
u
X
a
X
2 X
E   a   x   x   E 2
74
Social Mobility and Education in Britain
we include two such effects, labelled as INI and IN2. As shown in Table 4.2, in which inheritance effects are included along with hierarchy effects, INI applies in all diagonal cells and is intended to capture a general propensity for class inheritance, while IN2 applies, additionally, in the cells indicating immobility in Class 1 and in Class 4. In the case of Class 4, that of small employers and own-account workers, the possibility clearly exists of 'going concerns' or amounts of capital being passed on from parents to children. In the case of Class 1, while this largely comprises (higher-level) salaried managers and professionals, it also includes, as was earlier noted, a small number of large employers and independent professionals and also a probably larger number of managers and professionals whose employment status is somewhat ambiguous in that, as well as receiving a salary, they participate to some extent in business or practice profits. Thus, the possibility again arises of class immobility being maintained via direct inheritance and also perhaps through privileged intergenerational access to high level positions.
Status Affinity Effects. We explained in Chapter i that while our focus is on class and class mobility, we recognise social status as the basis of a further form of stratification that we may need at some points to take into account. The main line of status division in British society has for long been, and still remains, that between what might be called the 'white-collar' and 'blue-collar' worlds. This can be shown to be the main division that runs through the structure of more intimate social relations, such as close friendship and marriage or partnership,4 We thus introduce status affinity effects into our model as ones that promote class mobility in that they in part offset hierarchy effects insofar as mobility occurs only within either the white-collar or blue-collar worlds. Although the NS-SEC classes do not map perfectly on to status divisions, Classes 1,2 and 3 can be regarded as being very largely within the white-collar world, and Classes 5,6 and 7 very largely within the blue-collar world. Correspondingly, as shown in Table 4.3, which includes status affinity effects along with hierarchy and inheritance effects, a white-collar status affinity effect, AF1, is taken to apply in all cells indicating mobility between any two of the former three classes, and a blue-collar status affinity effect, AF2, in all cells indicating mobility between any two of the latter three classes. What this means is, for example, that under our model we would
4 This is evident from the status scale, based on the occupational structure of close friendship, developed by Chan and Goldthorpe (2004).
■3 -a
-a
K ti r-,
■s
k.
8
C
■f
5
-55
■a
H
U
c
' c o
■■-1—
a
GO
J3 U
u
u
u
X
JO
X
X
X X
(SI
X
X
X
ro ■—I
X
>—i
X
-I X
r Z
X
X
S
Z
*—<
X
x 2
x   x   x   5 5c
r-i
X
Z ?2
X
		rl	
IN	HI	HI	HI
	i—1 rH		T-H
s	X X	X	X
X
r-J
X
X
X
— X
rl
X
u
u
u
u
u
The Pattern of Social Fluidity within the Class Structure
77
X
o
c o
u
_3
G
u
u
u
HI2	rN X				s
X	X			g	5
HI2					X
T—1	—			T—1	
S3	X		g	5	£
HI2		IN2			HI2
HIl	HIl	INI		HIl	
S 6
§ £
SI
jz;     a     a     a     a     a a
AFl	cN X	2	HI2
HIl	HIl	IIH	X
is			
u	Clas!	Clas;	Clas:
IN
X
m X
X
Ü
rl
X
X
(N
X
-2
U
expect that although Class 3 and Class 5 are placed at the same hierarchical level, mobility between Class 3 and Classes 1 or 2 will be more likely, because of white-collar affinity effects, than mobility between Class
5 and Classes 1 or 2; and that mobility between Class 5 and Classes
6 or 7 will be more likely, because of blue-collar affinity effects, than mobility between Class 3 and Classes 6 or 7.
Jn sum, our model states that the pattern of social fluidity within the British class structure can be captured by the interplay of the class hierarchy, class inheritance and status affinity effects that it comprises -eight effects in all - as these operate in the cells of mobility tables based on the NS-SEC classes. That is to say, the numbers of individuals found in each cell will be determined by these effects, net of the effects of class structural change, and so too therefore will be all the odds ratios in terms of which social fluidity is defined. From Table 4.3, in which the distribution of the effects appears in full, it may be observed that there are six cells - those referring to mobility occurring between any two of the three intermediate classes, Classes 3, 4 and 5 - where none of the effects we distinguish are included. This is because the theoretical ideas underlying our model give no reason to do so, and we therefore suppose that the numbers in these cells will be determined simply in consequence of the effects operating in other cells. How well, we may now ask, does the model succeed in reproducing our empirical data?5
To begin with, we can apply the model to mobility tables for men in our 1946, 1958 and 1970 birth cohorts where class destinations can be determined at age 38 - that is, to the same tables we used in our analysis of relative rates of mobility in the previous chapter. If, first of all, we pool the data for all three cohorts, we find that the model does in fact fit the data well by standard tests. Only around 4 per cent of all individual cases are misclassified, and this lack of fit cannot be regarded as statistically significant. Moreover, if we then apply the model with the parameters for the eight effects being allowed to vary from cohort to cohort, we do not in this way achieve any significant improvement in fit. In other words, what is indicated is that, so far as men are concerned, not only is the level of fluidity highly stable over time, as was shown in Chapter 3, but likewise the pattern of fluidity. Over the historical period covered by our birth cohorts, the effects of class hierarchy, class
Full details of the model fitting leading to the results reported in the following paragraphs are provided in Bukodi, Goldthorpe and Kuha (2017).
78
Social Mobility and Education in Britain
inheritance and status affinity would appear to have operated in essentially constant ways in creating this pattern.
Next, we can apply the model to the corresponding mobility tables for women, although limited to those women who have worked only full-time and for whom, as for men, we find no evidence of change over time in the level of fluidity. The results that emerge are in fact the same as for men. With the data pooled across the cohorts, the model fits well, with again only around 4 per cent of all cases being misclassified, and again allowing the effect parameters to vary by cohort produces no significant improvement in fit.
However, while the overall pattern of fluidity can then be regarded as showing long-term stability for men and for women 'full-timers' alike, the further question remains of whether gender differences exist as regards the strength of the several effects that are involved. To investigate this possibility, we start by fitting our model to the mobility data pooled across both cohorts and genders. Once more we obtain a satisfactory fit, pointing to a large degree of commonality in the strength of the effects that prevail. Nonetheless, we do achieve an improvement in fit, of a slight but still significant kind, if we go on to allow the effect parameters to vary by gender. Further light can be thrown on the differences that arise if, as in Figure 4.1, the strengths of the different effect parameters are compared when the model is fitted to men and women separately.
In interpreting Figure 4.1, the following points should be kept in mind. First, hierarchy effects, in limiting mobility, are negative in sign because they depress the numbers in any off-diagonal cell of the mobility table in which they apply. Second, inheritance effects, in promoting immobility, are positive because they raise the numbers in the diagonal cells in which they apply. And, third, status affinity effects are also positive in that they serve to raise mobility by offsetting the negative hierarchy effects in those off-diagonal cells that relate to mobility between classes within the white-collar or blue-collar worlds. What then substantively emerges from Figure 4.1 is that gender differences sufficiently marked to attain statistical significance occur in only two respects, both of which are, however, of some interest.
The largest difference arises with the white-collar status affinity effect, AF1, which is clearly stronger for men than for women. In Figure 4.1 the length of the bars involved indicate that with men this effect is sufficiently large to offset both the HI'l and HI2 effects, while with women it offsets only the HI1 effect. In accounting for this difference, it is relevant to note that there is a tendency for women, even if from more advantaged class
The Pattern of Social Fluidity within the Class Structure
79
0.0 0.5 1.0 1.5
Effects'3' ■ Men        □ Women
Note
(a) These effects can be understood as those that, in combination, determine the numbers of individuals found in each cell of the 7 x 7 class mobility table to which they apply (Table 4.3)
Source: Bukodi et al. (2017)
Figure 4.1 Hierarchy, inheritance and status affinity effects, men and women 'full-rimers'
backgrounds, to be more concentrated than their male counterparts in employment in the lowest white-collar status groups - mainly those of routine administrative and clerical workers - and, further, that women are more likely than men to remain in such employment over the course of their working lives (see further Chapter 7).
A somewhat smaller difference is found with the general inheritance effect, INI, which is also stronger for men than for women. In accounting for this greater propensity for class immobility among men, what would appear chiefly important is that the tendency for men to follow their fathers in specific occupations is stronger than the tendency for women to follow their fathers - or indeed their mothers - in this way. In other words, one could say that men appear, whether for better or for worse, to be more caught up than women in family occupational traditions.6
By applying our model, we have now answered the first three of the questions from which we started out. So far as men and also women 'full-timers' are concerned, the pattern of fluidity within the
s This gender difference in the tendency for 'occupational succession' is evident in cross-national studies {see Jonsson et al., 2009; Erikson, Goldthorpe and Hallsten, 2012), and can be clearly shown with our own birth cohort data if we move down to consider mobility at an occupational rather than a class level.
80
Social Mobility and Education in Britain
class structure is one that can be adequately captured by class hierarchy, class inheritance and status affinity effects as these apply under our model; for both men and women this pattern proves to be highly stable over the historical period that we cover; and, while certain gender differences can be identified in the strength of effects, what is most notable is the degree of cross-gender commonality that is found. We can then move on to the last question we posed: that of what is implied for the chances of individuals making particular mobility transitions and for inequalities in these chances.
Since our model of social fluidity reproduces to a close approximation the numbers of individuals found in each of the cells of the seven-class mobility tables that we have constructed for men and for women full-timers in the 1946, 1958 and 1970 birth cohorts, it will likewise reproduce all the odds ratios that can be derived from these numbers. There are, it may be recalled, 441 odds ratios implicit in each of these mobility tables - one for every possible pair of the seven origin classes taken together with every possible pair of the seven destination classes. However, for our present purposes, we can concentrate on a particular, quite limited, subset of these odds ratios that is sociologically most readily interpretable and informative: that is, the subset of symmetrical odds ratios. An odds ratio is symmetrical where the pair of origin classes involved is the same as the pair of destination classes. Thus, an odds ratio giving the chances of an individual originating in Class 1 being found in Class 1 rather than in Class 2 relative to the chances of an individual originating in Class 2 being found in Class 1 rather than in Class 2 is symmetrical, whereas an odds ratio giving the relative chances of individuals originating in Class 1 and in Class 2 being found in, say, Class 5 rather than in Class 7 would not be symmetrical.
Figure 4.2 shows in graphical form the ranges of magnitude of the symmetrical odds ratios - twenty-one in all - that are implied by our seven-class mobility tables. Men and women are treated separately but in each case the data for the three cohorts are pooled - since we know that no significant change occurs across the cohorts. The first entry in the first row of the figure is the symmetrical odds ratio for Class 1 and Class 2, the next entry, that for Class 1 and Class 3 and so on down to the one entry in the last row which is the symmetrical odds ratio for Class 6 and Class 7. The ranges of the magnitudes of the ratios, as derived from our model, are indicated by the depth of shading of the blocks; the deeper the shading, the larger the ratio or, that is, the greater the inequality in the
The Pattern of Social Fluidity within the Class Structure
81
Class 2   Class 3   Class 4   Class 5   Class 6   Class 7
Class 1 Men
Women
Class 2 Men
Women
Class 3 Men
Women
Class 4 Men
Women
Class 5 Men
Women
Class 6 Men
Women
Range of odds ratios
< 2
I 10 < 20
Source: Bukodi et al. (2017)
Figure 4.2 Symmetrical odds ratios under the topological model for men and women 'full-timers'
relative mobility chances that are involved.7 If we think of Figure 4.2 as, so to speak, mapping a mobility terrain, with the shadings indicating contours, then the following three features are most notable.
First, cells in the top-left and bottom-right corners of the figure form rather 'flat' areas where inequalities in relative mobility chances are quite low, with odds ratios falling for the most part somewhere between 1, which would imply perfect mobility or the absence of any association between class origins and destinations, and 2. These are areas relating to
7 It may be asked why we derive the values of the symmetrical odds ratios from cell values under our model rather than from the values that we actually observe in our mobility tables. The answer is that where a theoretically informed model closely reproduces the empirical data, the values given under the model can be regarded as preferable to those calculated directly from the data in that the latter are likely to contain purely chances perturbations. And of course insofar as the model is a well-fitting one, the differences that arise will in any case be very slight.
82
Social Mobility and Education in Britain
mobility that falls within either the white-collar world of Classes 1,2 and 3 or the blue-collar world of Classes 5, 6 and 7. Thus, for example, one could say that the chances of someone originating in Class 1, the higher salariat, being found in Class 1 rather than in Class 2, the lower salariat, are not more than twice as great as the same chances for someone originating in Class 2, and an analogous statement could be made as regards the relative chances of mobility between Class 6 and Class 7, the higher and lower strata of the working class. The high fluidity that prevails in the cases in question results from such hierarchical barriers as arise being in large part offset by the white-collar and blue-collar status affinity effects that are included in our model (see Tables 4.1 and 4.3), and the - rather small - departures from perfect mobility that still occur are brought about primarily by the general class inheritance effect.
Second, in cells in the middle areas of Figure 4.2, relating to all mobility transitions occurring between origins and destinations in Class 2, in the three intermediate classes, Classes 3, 4, 5, and in Class 6, the inequalities become steeper, with the symmetrical odds ratios now in the main falling within the range of 2 to 5. This reduced fluidity derives from two main sources: mobility between, on the one hand, Classes 2 and 3, and on the other, Classes 5 and 6, entails crossing the division between the white- and blue-collar worlds so that hierarchy effects are no longer offset by status affinity effects; and, further, all mobility transitions involving Class 4, that of small employers and own account workers, are affected by the increased propensity for intergenerational immobility in this class that is captured by the additional inheritance effect (see Table 4.2).8
Third, in the cells relating to mobility transitions involving Class 1 and Class 7 that run along the top and down the right-hand side of Figure 4.2, apart from those in the flat corners previously discussed, inequalities rise again and sharply so. This is the result of successive hierarchical effects coming into play and also the additional inheritance effect that applies with Class 1 as well as with Class 4 (see, again, Tables 4.1 and 4.2). The symmetrical odds ratios indicating inequalities in relative chances of mobility between Class 1 and Classes 4 or 5 and between Classes 2 or 4 and Class 7 lie, with one exception, in the
8 As earlier noted, it is a shortcoming of the studies of income mobility that have been made in Britain that they leave small employers and the self-employed out of account because of the unreliability of their reporting of their incomes. But individuals falling into NS-SEC Class 4 now account for between 10 and IS per cent of the active population, and of late self-employment would appear to be steadily on the increase, even discounting its more bogus forms.
The Pattern of Social Fluidity within the Class Structure
83
range of 5 to 10, while those indicating such inequalities between Class 1 and Classes 6 or 7 are still higher, in the range of 10 to 20. For men, the ratio for mobility between Class 1 and Class 7 works out in fact at exactly 20: that is to say, the chances of a man originating in Class 1 being found in Class 1 rather than in Class 7 are estimated as being twenty times greater than the chances of a man originating in Class 7 being found in Class 1 rather than in Class 7. Mobility transitions made between the classes in question here are very much 'against the odds'.9
Upward mobility against the odds
Harold
Harold's father was a bricklayer, working on building sites, and his mother an office cleaner, so as he was growing up 'there wasn't a lot of money around'. However, his parents were quite strict and placed a strong emphasis on good manners and 'respectability'. They could not afford holidays, and, when a child, Harold spent his summers helping out on his grandfather's smallholding.
Harold was always 'quite academic' and won a place at a grammar school. He did well in all examinations and went on to university where he obtained a good first degree in physics. He then moved to another university to take a Ph.D and, on completing this, was offered a research position in a physics laboratory. However, he decided to become a science teacher.
He has made a very successful professional career, becoming Head of Science in a high-ranking school, and has the possibility of a school headship now open to him. His wife is also a teacher. They have known each other since their schooldays together and, Harold says, 'she is really, really my very best friend' and his main source of support. They have no children and so live in an apartment in a pleasant suburb of a large city. But they are 'Mediterranean enthusiasts', for the sun and food, spend all their holidays in a villa on the Italian coast, and think that they may eventually retire there.
The inequalities in relative mobility chances reported here are clearly greater than those found in previous studies covering in part at least the same historical period (e.g. Goldthorpe and Jackson, 2007; Li and Devine, 2011). This would appear to be the result of NS-SEC providing a more accurate and reliable instrument for determining class positions than the classifications previously in use. See further Bukodi et al. (2015).
.....
84
Social Mobility and Education in Brita
in
Upward mobility against the odds - 2
Gordon
Gordon grew up in a large working-class family living on the top floor of a tenement building: 'lino on the floors and coal fires'. He did well at school, always being top of his class, and got good O levels in a range of 'applied' subjects. But when he was 16 his father died so Gordon, who says that by this time he had become 'a bit of a tear-away' and a member of a street gang, decided to leave school to earn some money. He started training as a surveyor, but this did not work out well, and he switched to an apprenticeship in carpentry.
On completing his apprenticeship, Gordon married, and he and his wife decided to join with an older brother who had set up his own construction firm. Gordon took on managerial responsibilities from the start. The firm became highly successful until, following some misfortunes in the late 1980s, it failed and went into receivership. However, after a few years, Gordon and his brother rebuilt the business and, after his brother died, Gordon took over full control, again with his wife as his main assistant: 'She has been my best friend, my pal, ever since we met ... We are together 24/7.'
The firm is now in good shape, with some thirty employees on the books and a steady flow of work coming in. Gordon and his wife have been able to send their children to private schools and have a large house in a 'village-like' suburb. They work hard bur go away on 'weekend breaks' every month.
As might be expected from what has been said earlier, the gender differences that show up in Figure 4.2 are not great. There is, overall, a tendency for symmetrical odds ratios to be lower for women than for men, which is chiefly because, as was noted, the general class inheritance effect is weaker for women. But the particular differences that arise in this way are sufficiently large to reach statistical significance in only three cases: that is, with the ratios indicating relative mobility chances between Class 2 and Class 5 and between Class 3 and Classes 5 and 6. As Figure 4.2 shows, for women these ratios all fall into the lowest range that we distinguish. There is one contrary instance where the ratio for women is higher than that for men. This occurs with relative mobility chances between Class 1 and Class 3, resulting, again
The Pattern of Social Fluidity within the Class Structure
85
aS was previously discussed, from women benefiting less than men from the white-collar status affinity effect so that hierarchical barriers to mobility within the white-collar world are stronger in their case. At the extremes, gender differences are slight.
Upward mobility against the odds - 3
Carol
Carol's father was a bus driver. Her family was always short of money, even for food. Her father gave her brothers encouragement in their education but less so Carol: 'he believed that a woman's place was in the home'. Carol was in fact doing well enough in secondary school to be a university applicant but 'rebelled' and left at age 17. Between then and age 26, while in a series of routine jobs, she had two failed marriages and was left as a single mother with two children. She also agreed to be foster parent to a relative's child, since she enjoyed 'mothering'.
In this situation Carol decided to try to resume her education and get to university as a mature student. She was accepted but found that she was not eligible for any grants. She therefore took a degree in social work on a part-time basis, while continuing to look after her own children and doing more fostering in order to support herself financially. She says of her education: 'I did it all arse upwards.'
After completing her degree, she became a social worker, then moved into social work management, and eventually became a senior manager in charge of children's services for a large regional authority. She now works independently as a consultant. As well as having gained 'complete financial security', she also has a more stable personal life with a third husband. They feel they will both be able to retire at 55. They presently live in a 'gated' community but are having a new house built for them in the countryside. They enjoy golf, skiing and travelling the world on holiday.
What, then, are the implications of our findings on the pattern of fluidity existing within the British class structure when set in relation to current political discussion of social mobility?
An initial point to be made is that our finding of an essential stability in the pattern as well as in the level of social fluidity across a period of more than half a century clearly reinforces the idea we introduced at
86
Social Mobility and Education in Brit,
am
the end of the last chapter: that of an endogenous mobility regime that is powerfully resistant to change. The effects on social fluidity of class hierarchy, class inheritance and status affinities persist over time in what can only be regarded as a rather remarkable fashion. We can therefore reassert that the abiding concern in political circles with change in social mobility - that is, with a supposed decline - is misdirected. The focus of attention should be not on change, in any direction, hut, to the contrary, on its absence.
Further, the results we have reported call into question two, not entirely consistent, claims that are often made or at least implied: first, that relative mobility chances in Britain are in all respects of a very unequal kind; and second, that a quite distinctive problem arises with inequalities in access to certain elite groupings - as mediated, say, through the influence of exclusive schools, 'Oxbridge' and metropolitan social networks.10 The conclusions to which our findings would point are, first, that some very wide variation exists in the degree of inequality of chances that are involved in different mobility transitions; but, second, that this variation is, in a rather systematic way, continuous rather than discontinuous.
On the one hand, our model of the prevailing pattern of social fluidity reveals that as regards certain mobility transitions a situation not widely divergent from that of perfect mobility does in fact prevail: that is, most importantly, in the case of mobility within the white-collar world of Classes 1, 2 and 3 - especially for men - and likewise within the blue-collar world of Classes 5, 6 and 7. The barriers to mobility in these regions of the class structure are not high, and it is in this connection relevant to recall that, as shown in Chapter 2, the total, absolute mobility rate in Britain, based on the seven NS-SEC classes, appears stable over time at around 80 per cent: that is, on this basis four out of five
10 The Social Mobility Commission, for example, in its annual reports has tended to take a generally undifferentiated view of inequalities in mobility chances among the population at large, while at the same time apparently regarding elite mobility as a special case (see e.g. Social Mobility and Child Poverty Commission, 2014). In the academic context, sociologists associated with the Great British Class Survey have argued that the data from this survey, even if unsuitable for the study of mobility at a population level because of its lack of representativeness, can still be a reliable source for treating what are taken to be largely separate but now crucial issues of elite mobility. See Savage (2015), but also, the cogent critique by Mills (2015) of both the conceptual and empirical bases of the studies undertaken.
The Tattern of Social Fluidity within the Class Structure
87
"ndfviduals are found in adult life in a different class to that in which they originated. To be sure, much of the mobility in question here is of only a short-range kind, but this is not to say that it should be discounted. In the light of the evidence on the extent of class inequalities that was presented in Chapter 1, mobility, whether upward or downward, between, say, Class 1 and Class 3 or Class 5 and Class 7 could scarcely be regarded as inconsequential. Our results in this regard do, incidentally, undermine suggestions that odds ratios are a statistic favoured by sociologists who are ideologically committed to the view of extreme inequality of opportunity in British society - because the quite unrealistic base of 1 serves to make all actual odds ratios appear excessive.11 However, for some mobility transitions odds ratios not far removed from 1 can in fact be shown to occur, so that comparisons made with the higher odds ratios arising with other transitions are entirely appropriate.
On the other hand, as the range of mobility extends so that the white-collar-blue-collar division is crossed and hierarchy effects come increasingly into play, the odds ratios that emerge from our model do indicate that inequalities in relative mobility chances progressively widen to a rather extreme degree. And in view of this, it becomes open to some doubt whether a focus on elite groupings - unless of a very specialised and minoritarian kind - is likely to reveal inequalities that are of a quite different order of magnitude. At all events, recent research on elite mobility has not produced results that would lend any very compelling support for this possibility. The important point may be that the inequalities in mobility chances that arise in the case of elite groupings do not imply a step change from those indicated by the odds ratios reaching up to 20 that we have reported - and that refer to mobility between classes covering quite substantial sections of the population. At the present time Class 1 accounts for around 10 per cent of all economically active individuals and Class 7 for upwards of 15 per cent.12 A preoccupation with inequalities in access to elites may
" See, for example, Saunders (2010: 26-32) and Payne (2017: 177).
12 The Social Mobility and Child Poverty Commission's study of elite mobility in Britain (2014) covered only a limited range of - somewhat arbitrarily defined -elites, analysed recruitment only in terms of school and university attended, with no reference to class origins, and made no attempt to determine statistically the extent to which the elites considered were more exclusive than higher level managerial and professional occupations in general. Reeves et al. (2017), focusing on individuals included in Who's Who - approximately 0.05 per cent of the total adult population - also have no data on class origins but do calculate
88
Social Mobility and Education in Brttain I The Pattern of Social Fluidity within the Class Structure
89
therefore lead to an undue disregard of restrictions on intergenera-tional class mobility that, while perhaps not all that less severe, are far more extensive, and that could, for this reason, be regarded as being, if anything, of greater concern from the normative standpoint of equality of opportunity.13
Finally, and most importantly, the results emerging from Figure 4.2 point to serious difficulties with the widely accepted political argument, or at least assumption, that promoting social mobility is the most effective response to the present-day problem of increasing inequalities of condition. Through seeking to increase equality of opportunity, and thus social mobility, it is supposed, inequalities of condition can be given greater meritocratic legitimation. However, what is indicated by our analyses of the pattern of relative mobility rates is that this strategy is highly questionable in that equality of opportunity, and its expression via social mobility, appears to be systematically compromised by inequalities of condition. The class hierarchy and inheritance effects that are included in our model serve, in the ways explained at the start of the chapter, to capture such inequalities. And what the model then shows is that it is where these effects are most limited that relative mobility chances are most equal, and social fluidity is at its highest level, while as these effects come into fuller operation, relative mobility chances become increasingly disparate - to a point at which any conception of equality of opportunity becomes difficult to sustain.
some odds ratios in relation to type of school attended. Thus, for the period 2001 to 2016, men and women who had been at Headmasters' and Headmistresses1 Conference schools - which account for around 2.5 per cent of all children in secondary education - were, as compared to others, thirty-five times more likely to become included, rather than not included, in Who's Who. While no direct comparison can of course be made with our odds ratios relating class origins to class destinations, no sharp discontinuity in chances is suggested so far as mobility between Classes 1 and 7 is concerned, and especially in view of evidence presented by Laurison and Friedman (2016) indicating that odds ratios referring to mobility chances between only the 'traditional' professions within Class 1 and Class 7 could go well above the 20 mark.
Reeves (2017) is a forceful critique of the preoccupation in current American social and political commentary with the most advantaged 0.1 or t.O per cent of the population, to the neglect of what would appear to be a steadily widening gap between the 'upper middle class' - defined as some 20 per cent of all Americans - and the rest of the population in terms of material living standards, quality of life and intergenerational mobility chances.
It is in this regard that the disconnect that exists between the discussion of social mobility in political and related policy contexts and what has been learnt from sociological research is perhaps most fundamental. In tne numerous reports and official or quasi-official policy documents dealing with mobility that we referred to in the Introduction, very little mention can be found of the possibility that, rather than greater equality of opportunity being a means of offsetting gteater inequality of condition, it will only be achieved if the reduction of the lattet is itself a primary policy objective. This issue becomes of particu-lat importance in regard to socially grounded inequalities in educational attainment. For reducing such educational inequalities is seen as crucial to increasing equality of opportunity and thus levels of mobility - as being in effect the key way in which the link between inequality of condition and inequality of opportunity can be broken. In the chapters that follow we turn our attention to the role education has actually played in mobility, and specifically in class mobility, in British society over the decades since the Second World War; and we ask how far here too a gap exists between the assumptions and beliefs that are built into prevailing political and policy discourse and the evidence of sociological research.