7 Education and the Labour Market: Is Education Now Class Destiny? In Chapter 5 we noted that theorists of postindustrial society, writing in the later twentieth century from a broadly liberal position, claimed that a movement towards an education-based meritocracy was in progress. At the same time as inequalities in educational opportunity were being reduced, educational attainment was becoming increasingly important in determining individuals' economic futures - their life-chances in the labour market. In this latter connection, the argument was developed along the following lines. In postindustrial society, characterised by rapid technological and economic advance, theoretical - as opposed to merely empirical - knowledge becomes of centra! importance. Thus, educational institutions, as the disseminators of theoretical knowledge and of associated expertise and skills, must take on a dominant role in the allocation of individuals to different positions within the division of labour. In particular, the university no longer serves, as in the past, chiefly to reproduce the status order of society but becomes, rather, 'the arbiter of class position' and in this way 'gains quasi-monopoly power over the fate of individuals'. Access to more advantaged class positions will lie beyond the reach of men and women, whatever their social origins and whatever their basic ability, if they lack appropriate qualifications; and a university degree serves as the essential 'passport of recognition'. This is the case, moreover, not only with professional and technical positions. With managerial and higher administrative positions also, the growing requirement for theoretical knowledge means that 'promotion from below' - as, say, on the basis of work experience and level of performance - which was once common, now becomes increasingly rare. In other words, it is achievement within hjucation and the labour Market 129 \v educational system, prior to entry into the labour market, that k ever more, class destiny.1 The influence of such ideas remains pervasive. They are, implicitly if not explicitly, embedded in much present-day political discussion of education and social mobility. For example, references to the emerging •knowledge economy' and its educational demands are frequent; formal qualifications are seen as ever more necessary for success in the labour market; and in turn educational policy, focused on expansion and the widening of opportunity, is regarded as vital in enhancing national economic performance and promoting social mobility alike. Nonetheless, the factual question remains of how far it is the case that ,n British society education is now class destiny or is, at all events, becoming more so. Remarkably little evidence of a directly relevant kind has thus far been produced. In seeking to take matters further in this regard, we have to develop our approach from that of previous chapters. Hitherto, we have analysed data from the three successive birth cohort studies on which we concentrate in order, primarily, to gain an understanding of change -or constancy - over historical time. Now, in addition, we need to exploit more fully the possibility we noted in the Introduction of using the rich data provided by these studies in order to consider change or constancy over individuals' life courses. Insofar as education is becoming class destiny, we should find that there is a tendency across our cohorts for individuals' class positions to become more stable within their working lives: that is, a tendency for the class positions they obtain on first entering the labour market to become increasingly ones in which they subsequently remain. And, in particular, the probability of their achieving upward class mobility over the course of their working lives should fall away. First of all, therefore, we aim to establish typologies of the 'class trajectories' that have in fact been followed by men and women in the 1946, 1958 and 1970 birth cohorts from the time of their labour market entry up to mid-life, so that, on the basis of these typologies, the extent of any changes in class trajectories occurring across the cohorts can be determined. Then, secondly, we move on to examine 1 As we previously observed (Chapter 5, o. 2), the best-known theonst of postindustrial society is Daniel Bell, and the phrases quoted in the text are all taken from his work (Bell, 1972: 30-1; Bell, 1973: ch. 30). 130 Social Mobility and EducaÜ how far and in what ways individuals' type of class trajectory js J to their level of education at labour market entry, and how far th^' indications that an education-based meritocracy is emerging i" regard, however, we need also to consider the further associate ' independently exist between individuals' class trajectories and one hand, their class of origin and, on the other hand, their co ability. For if an education-based meritocracy is coming into these latter associations should be weakening. Class origins and< tive ability should be of declining importance for individuals' trajectories except insofar as their effects may be mediated thro educational attainment. For the majority of men and women in each birth cohort we data that allow us to reconstruct, more or less completely, employment and occupational histories from labour market entry | to age 38, an age by which, as we earlier noted, any occupation, changes that an individual may make become increasingly less likelyt entail a change in class position. Thus, for each of these coh members there is the possibility of allocating them, year by yeari a period of twenty-two years - from age I 6 to age 38 - to one orothtr of the classes of the seven-class version of NS-SFC (Table 1.1), referring for this purpose to their 'modal' class in each year - that in whitk they had spent most time. In addition, though, we have to recognise two other possibilities for each year: first, that the cohort member had not yet entered the labour market; and, second, that for that year their class position is unknown - information is missing.2 We have, therefore, for each cohort member a total of nine different 'states' in which he or she may be found in each one of a sequence of twenty-two years. This is a unique body of data, and since for the three cohorts together In cases where for whatever reason - including, as often with women, their absence from the labour market - we have no information on a cohort membert class position for ten consecutive years, they are dropped from the analysis. We also, of necessity, exclude cohort members who have never been in employment, and further those whose first significant job - i.e. one lasting for six months of more - occurred only after age 30. The proportion of original cohort member* excluded for one or other of these reasons is only around 6 per cent in the 1970 cohort but close to 20 per cent in the 1946 and 1958 cohorts. I lowever.dS would appear to be the result in these two earlier cohorts largely of operational shortcomings in data collection and coding procedures rather than of cohort members' non-response, and thus any associated biases should be minimised (Nathan, 1999). Ilioii hau' sue to tuque The basis of this technique is the pairwise matching of each qUejCeS| sequence with each other individual sequence, in the sense of Id the Labour Market 131 •h sequences for 12,886 men and 11,715 women, the data task that arises is a demanding one. ^r°CeS:>' hie us ro moVL"*rum t'u' Sundance of data we have available 1 nstruction of typologies of class trajectories, we use a tech-ence analysis known as 'optimal matching' - initially n molecular biology in order to study complex protein < E CD * i in *- co CO CO o) TO w I. >s O c ■— CO 1^ co £ O CO 8 & 55 + CD V) O II o ^ -i- CO II m "D c CD CD O C C cr CD (75 TO CD 0) O c CD a- CD T3 S= CD CN CD O c CD a" CD CO 0 (N c 'a. CD 1 "a c co CD Q. i_ O o O c u T a 5 u CD O O 00 3 the same as Sequence 1 with the single exception that at around age 24 rhis individual - say, as a result of a career shift - spent a year in a Class 2 position before returning to Class 1 positions. In this case, notching the two sequences is very easy and 'low cost': for just one ear one substitution is needed - Class 2 for Class 1 - and this involves only one 'eve' m t'ie c'ass h'erarcnv- Thus, following our costing system, the distance between Sequence 1 and Sequence 2 will be determined as very small, and is in fact, at 1, at the minimum possible level. In contrast, consider matching Sequence 1 - or Sequence 2 - with Sequence 3. In this latter case, an individual entered the labour market at Jge 16 in a Class 7 position and remained continuously in this or in other Class 7 positions through to age 38. That is to say, Sequence 3 differs from Sequences 1 in every one of the twenty-three years covered, and so far as class states are concerned the differences are as wide as could be within the class hierarchy. Thus, the distance between Sequence 3 and Sequence 1 will be determined as very large, and is, at ]00, close to the maximum possible level under our costing system -while that between Sequence 3 and Sequence 2 is almost as large at 99. We should, however, emphasise that the distances between the actual sequences followed by members of our three cohorts are very variable, lying on all intermediate points within the range indicated by the examples of Figure 7.1. Indeed, the really striking feature of the sequences is the enormous diversity that class histories display when considered at a detailed level. On the basis of our costing system we can then apply an optimal matching algorithm in order to produce matrices - for men and women separately - that give, pairwise, the distances between the sequence followed by each man or woman in our cohorts and that followed by every other man or woman. As an analogy, think of the kind of chart found in a road atlas that shows the distance between each town or city and every other town or city. These distance matrices represent the initial step in the process of bringing the vast amount of data that we have on class histories into some manageable form. But - unlike road atlas charts - they are in themselves far too large to be readily inter-pretable. We are dealing with over 85 million pairs of distances for men and with over 68 million for women. Our next step, therefore, following the usual practice in optimal matching, is to move to a process of 'cluster analysis' - that is, one that involves searching the distance matrices for clusters of more or less similar sequences - and it 134 Social Mobility and Education in ß "toil is then the clusters thus found that serve as the basis for our typologjjj of class trajectories.4 In the light of our cluster analyses, we in fact opt for an eight-cluster typology of class trajectories for men and a nine-cluster typology f0r women. In Tables 7.1 and 7.2 we describe the types of class trajectory that emerge and for each type give an illustrative sequence in the same form as in Figure 7.1. But we would again stress that, because of the great diversity of the actual trajectories followed, very many variants on these illustrative sequences occur, even if ones that for the most pa^ reflect short-distance shifts within the class hierarchy over shon periods of time. We are dealing here with empirically derived, not pure or 'ideal', types. As can be seen from Tables 7.1 and 7.2, the typologies for men and women are broadly similar. In both cases there are types of class trajectory representing stability: that is, stability within Classes 1 and 2, the managerial and professional salariat (Types 1 and 2 for men and women alike); within Class 3, ancillary professional and administrative employees (Type 4 for men and Type 5 for women); within Class 4, small employers and self-employed workers (Type 5 for men and Type 6 for women); and within Class 6, the upper stratum of the wage-earning working class (Type 7 for men and Type 8 for women). The main differences that arise are three. First, for women there is no type representing stability in Class 5, lower supervisory and technical employees, corresponding to Type 6 for men. Second, while for men there is only one 'upward mobility' type (Type 3), for women there are two (Types 3 and 4) - a distinction being made between trajectories where mobility tends to occur relatively early in working life, before around age 30, or only later. And third, while for both men and 4 We use what is known as 'Ward's method' of cluster analysis. This results in a treelike plot, or dendrogram, that moves from a large number of small, well-defined clusters (the leaves) through a smaller number of larger clusters (the branches) to a still smaller number of still larger but less well defined clusters (the trunks). The analyst has then to choose at what level the line is to be drawn through the dendrogram - usually somewhere among the branches - that will determine the number of clusters to be distinguished. We have checked the empirical quality of the clusters that we work with by using 'silhouette widths' which compare the distance of each individual sequence from its cluster centre with its distance from the centre of the next nearest cluster. In general, we achieve satisfactory results. The least well-defined clusters are those which become the Type 3 trajectory for men and the Types 4 and 9 trajectories for women. f , catm and the Labour Market 135 n there are types representing mobility within the working class Ivpes 8 and 9 respectively), only for women is there a type for lownwafd mobility to the working class (Type 7). In carrying out our optimal matching exercise, in order to arrive at these typologies of class trajectories, we have worked with the class histories of men and women in all three of our cohorts taken together. Now, however, in order to address the question of whether there has keen change in class trajectories over historical time - and in particular of whether an increasing stability in class positions from labour market entry onwards is apparent - we need to consider men and women in the three cohorts separately, and to ask how far types of class trajectory differ in their frequency from cohort to cohort.6 The relevant distributions are shown in Figure 7.2 for men and in Figure 7.3 for women. To begin with the distributions for men, it can be seen that there is a clear rise across the cohorts in the proportion following Type 1 class trajectories - those characterised by late entry into the labour market and then stability in the higher-level managerial and professional positions of Class 1. This finding is therefore consistent with the expectation that such positions should be increasingly ones that are directly entered into by individuals who have spent a lengthy prior period in full-time education - rather than being filled by promotion from below. However, it has also to be observed that there is no corresponding increase in the proportion of men following Type 2 trajectories - those characterised by late entry into and stability within the lower level managerial and professional positions of Class 2. Furthermore, while the proportion of men with Type 3 trajectories, involving upward worklife mobility, including into Classes 1 and 2, from labour market entry in less advantaged class positions does decline slightly between the 1946 and 1958 cohorts, this decline does not continue through to the 1970 cohort. And it is also of interest to find that there is no clear trend of change in the case of Type 5 trajectories that involve men being mobile from Classes 5, 6 and 7 positions to become small employers or self-employed workers in Class 4 - a form of possible worklife * The question may be raised of how far, if we had treated our cohorts separately in constructing our typologies, we would have ended up with the same typology in each case. We investigated this possibility and found that, with both men and women, differences across the cohorts were very small, and insufficient to abandon the obvious advantages of using the same typology for all three cohorts. Table 7.1 Typology of class trajectories, men TyPes Illustrative sequence1'1' (1) Majority: late entry and stability in Class 1 Minority: upward mobility to Class 1 from Classes 2 and 3, before age 30 Age 16 -1-1- p. o 1 0 [ 0 0 0 0 J 0 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Class 2 Minority: upward mobility to Class 2 from Class 3, before age 30 (3) Majority: upward mobility to Classes 1, 2 and 3 from Classes 4, 5, 6, 7, after age 30 Minority: downward mobility to Classes 4, 5, 6, 7 from Classes 1, 2, 3, after age 30 (4) Majority: stability in Class 3 Minority: upward mobility to Class 2 from Class 3, after age 30 in 0 1 --1- 0 J 0 0 0 0 3 | 3 2 i 2 0 0 0 6 6 6 5 5 5 J 5 5 5 5 5 5 3 |T" 3 3 2 2 2 2 3 3 3 3 3 3 ihhhhhhli] (5) Majority: mobility to Class 4 from Classes 5, 6, 7, before age 30 Minority: stability in Class 4 (6) Majority: stability in Class 5 Minority: upward mobility to Class 5 from Classes 6 and 7, before age 30 (7) Majority: stability in Class 6 Minority: mobility to Class 6 from Classes 5 and 7, before age 30 (8) Majority: mobility into and out of Class 7, after age 30 Minority: stability in Class 7 Note (a) Notation as in Figure 7.1 Source: Bukodi, Goldthorpe, Waller and Halpin (2016) 4 4 4 4 4 4^4\4\4^4\4\4\4\T -1-1 I -1—1 0 0 6 6 6 5 >|< S 5 5 1 5 1 5 L5 | 5 | 5 | 5 5|5|5l5 5\s\ 0 0 0 0 7 6 6 6 7 7 6 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 6 6 6 6 7 7 6 6 7 7 7 7 7 ; Table 7.2 Typology of class trajectories, women Types__Illustrative sequence'3' (1) Majority: late entry and stability in Class 1 Minority: upward mobility to Class 1 from Classes 2 and 3, before age 30 (2) Majority: late entry and stability in Class 2 Minority: late entry and stability in Class 3 (3) Majority: upward mobility to Class 2 from Classes 3 and 6, before age 30 Minority: stability in Class 2 (4) Majority: upward mobility to Classes 2, 3 and 5 from Classes 6 and 7, after age 30 Minority: mobility between Classes 4, 5, 6 and 7 1-1- Ape ^ 0 1 0 0 0 0 0 0 2 2 1 1 j 1 1 1 1 1 | 1 | 1 1 1 1 1 j 1 0 0 0 0 0 0 0 0 3 3 I 2 2 2 2 2 2 2 2 2 2 2 2 2 1 0 0 0 6 6 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 0 0 7 7 7 6 6 6 6 6 6 6 6 6 6 5 5 5 »1 3 3 2 2 (5) Majority, stability in Class 3 Minority: mobility to Class 3 from Class 6, before age 30 (6) Majority: mobility to Class 4 from Classes 3, 6, 7, before age 30 Minority: stability in Class 4 (7) Majority: downward mobility to Classes 6 and 7 from Class 3 0 3 3 3U\3\3\3\a\3\3\»\>\>\»\»T^Tra 0 6 3 3 4 4 4 4 4 4 4 4 4 0 3 3 3 3 3 6 7 7 7 (8) Majority: stability in Class 6 0 7 7 7 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Minority: upward mobility to Class 6 from Class 7, before age 30 (9) Majority: mobility between Class 0 0 0 0 7 7 7 7 7 6 6 6 7 7 7 7 6 6 6 6 6 7 7 6 and 7 Note (a) Notation as in Figure 7.1 Source: Bukodi, Goldthorpe, Waller and Halpin (2016) 140 Social Mobility and Education in Britain (a) Type 1 Type 3 Type 5 Type 7 25 20 15 10 5 0 25 20 15 10 5 0 25 20 15 10 5 0 25 20 15 10 5 0 Type 2 Type 4 Type 6 Type 8 1958 cohort 1970 Note (a) Types as characterised in Table 7.1 1958 cohort 1970 Source: Bukodi, Goldthorpe, Waller and Halpin (2016) Figure 7.2 Distribution (%) of class trajectory types across cohorts, men advancement that has also been seen as increasingly restricted in post-industrial society.7 Overall, then, what is most notable about the results shown in Figure 7.2 is the rather modest degree of difference in the distributions that shows up across the cohorts, given the historical period that they span. In particular, no clear tendency is apparent for class trajectories that imply some degree of upward worklife mobility to become any less frequent.8 7 See e.g. Bell (1972: 30) and for critical comment, Arum and Müller (2004). In recent years there has been a notable increase in essentially bogus self-employment where workers are constrained to take on self-employed status by those who ate in effect their employers. However, this increase is unlikely to have had much effect on the results we teport since these extend at the latest only to 2008: i.e. to men at age 38 in the 1970 cohort. 8 Similar results have in fact been earlier reported by Mills and Payne (1989), using a different dataset, that of the ESRC Social Change and Economic Life Initiative, and a different analytical approach. education and the Labour Market 141 (a) Type 1 30 25 20 £ä 15 10 5 0 Typo 3 30 26 20 15 10 5 0 Type 5 30 26 20 £ 15 10 5 0 Type 7 30 25 20 15 10 5 0 Type 2 Type 4 Type 6 Type fl 30 25 20 15 10 5 0 30 25 20 15 10 5 0 30 25 20 15 10 5 0 30 25 20 B« 15 10 5 0 1958 cohort 1970 TypeQ 30 25 20 3= 15 10 5 0 1946 1958 cohort 1946 1958 cohort 1970 (a) Types as characterised in Table 7-1 Source: Bukodi, Goldlhorpe, Waller and Halpin (2016) Figure 7.3 Distribution (%) of class trajectory types across cohorts, women Turning to the corresponding distributions for women, as shown in Figure 7.3, we find in several respects a similar situation as with men. Most importantly, there is again an increase in Type 1 class trajectories characterised by late entry into the labour market and then stability in Class 1 positions, although, again too, this increase is not replicated with Type 2 trajectories characterised by late entry and stability in Class 2. However, a difference arises in that while with men Type 3, upward worklife mobility, trajectories could be said at all events not to 142 Social Mobility and Education in Bn 'tail, be in steady decline, with women Type 3 trajectories, involving upward worklife mobility mainly from Class 6 and 7 to Class 2 positions and achieved before age 30 actually show an increase. And at the very least, no falling off is apparent in the case of Type 4 trajectories involving more limited upward mobility at a later stage in women's working lives. The growing importance of Type 1, 3 and 4 trajectories together is then chiefly offset by a quite substantial decline in the proportion following Type 9 trajectories, those characterised by mobility between Class 6 and 7 positions or, that is, within the wage-earning working class. So far, therefore, the results of our analyses are somewhat equivocal as regards the question of whether in Britain an education-based meritocracy is emerging, so that the class positions in which individuals first enter the labour market on completion of their education become increasingly determinative of their class histories. We have shown that both men and women tend more frequently to take up Class 1 positions on labour market entry and to remain in such positions subsequently; but we have found no clear evidence, in the case of men, of a decline in class trajectories entailing upward worklife mobility while, in the case of women, such trajectories would appear to be rising in frequency. In order to gain a more decisive outcome, we now move on to results from analyses in which we directly examine the extent to which different types of class trajectory are associated with individuals' educational attainment at labour market entry - but also, for the reasons indicated at the start of the chapter, with their class origins and their cognitive ability. As we have seen in Chapter 6, individuals' class origins and cognitive ability are both strongly associated with their educational attainment. But what, for present purposes, we further wish to know is how far they have additional, independent effects on individuals' class trajectories - that is, effects over and above those that operate via education. In Tables 7.3 and 7.4 we show, on the basis of the analyses we have undertaken, where these various possible effects on class trajectories of men and women are of positive or negative statistical significance, taking our three cohorts together, but with cohort being included in the analyses as a control variable. Since we are here primarily concerned with education as an investment good in relation to the labour market, we revert to treating educational attainment in terms of the four relative levels of qualification, the education and the Labour Market 143 co nstruction we explained in Chapter 5. To recall, the highest relative level includes all qualifications from A levels upwards for the 1^46 cohort, but is restricted to tertiary-level qualifications for the 1958 cohort, and further to degree-level qualifications for the 1970 cohort, while the lowest level is no qualifications for the 1946 cohort but extended to no qualifications or any sub-secondary qualification for the 1958 and 1970 cohorts. We wish to determine individuals' qualifications at the start of their working lives. With men and women in the 1958 and 1970 cohorts we take their qualifications at the time of their first significant job (one lasting for at least six months) but, because of data limitations, with men and women in the 1946 cohort we have to take their qualifications at age 26.9 To simplify the results somewhat, the NS SEC classes are collapsed in the way indicated in the tables, and the two highest and two lowest cognitive ability quintiles are also collapsed. Since the results we obtain are on much the same pattern for men and women, we discuss Tables 7.3 and 7.4 together. What may be observed, first of all, is that Type 1 and Type 2 class trajectories, characterised for men and women alike by stability within Class 1 and Class 2, are - as would be expected given the late entry into the labour market that is also involved - positively associated with the highest relative level of qualification. A clearly different situation is thus indicated from that in which, as revealed in the Nuffield mobility study of 1972, a large majority of those men born in the first half of the twentieth century who gained access to managerial and professional positions did so while having only very modest qualifications on leaving full-time education, if indeed, in the case of managers, any qualifications at all. And, conversely, it can be seen that those class trajectories that are largely restricted to positions within Classes 6 and 7 - that is, Types 7 and 8 for men and Types 8 and 9 for women - are positively associated with the two lowest qualification levels. In other words, for the period our data 9 Some overestimation of level of qualification at labour market entry may in this case arise, on account of qualifications acquired between entry and age 26, but this is likely to be slight. We know that at age 26 stil! around three-fifths of the members of the 1946 cohort had no more than lower-level secondary qualifications. The statistical model from which the results reported in Tables 7.1 and 7.2 derive is a multinomial logistic regression model, full details of which are given in Bukodi et a.l. (2016). 144 Table 73 Effects of relative level of qualificati class of origin and cognitive ability on t>rnhr,i°?- "*market types of class trajectory, men™ prob"bihty °f foU0uHn*^H Relative level of qualification lowest next to lowest next to highest highest ++ Class of origin Classes 6 and 7 Class 4 ns Classes 3 and 5 + Classes 1 and 2 + Cognitive ability quintile lowest two — middle highest two Class 3 Rectory type*) 4 5 ns refere + + nee MS + ++ + ns ns ns + reference ns ++ ns ns + ns Wote (a) For symbols, see Table 6.1 (b) As shown in Table 7.1 Source: Bukodi, Goldthorpe, Waller and Halpin (2016) cover, educational attainment has evident! ant part in determining whether individuate"'16 i° ^ 3n imP°n' within the managerial and Sf S ^stories -----------0 --------M individuals' cl - — un- Jargely spent within the managerial and f ■ lst°r,« are within the wage-earning working class To ^ eSSIOna' sa'ariat or the idea of an emerging education-based m ' ^ eXtent' therefore to find support. meritocracy would appear' what has further to be observed is rh are also positively associated with CltL i \ And Type However,--------------.------n served is fhar T 2 trajectories are also positively associated with CI Pe 1 and TYP' while Type 7 and 8 trajectories for men and Type 8 ^ ^ ^ 2 ori8'ns-negatively associated with such origins, even ^h ^ 9 ^ Wwrneri also included in the analysis. That is to say Co ^ ^catt°ns -lass origins has a significant effect on ,n^°T'n.SJrotn advantaged are are ncganveiy as>ui.idic also included in the analy^ class origins has a significant effect on indivicTT'advantaged independently of the qualifications that they hav * S C'aSS ^tes ^u'ie and, likewise independently of qualification* tJ* °f ^ave not' gained ns^hose trajectories that are Education and the Labour Market 145 Table 7.4 Effects of relative level of aualiRr ti class of origin and cognitive ability on pmb hT * types of class trajectory, women(a> ability of following different Relative level of qualification lowest next to lowest next to highest highest Class of origin Classes 6 and 7 Class 4 Classes 3 and 5 Classes 1 and 2 Cognitive ability quintile lowest two middle highest two _C^_trajectory type'1" 6 7 ns + ns ns ns + ns t + reference ns ns ns + ~ ns + ns + ns ns ns ns ns Notes (a) For symbols, see Table 6.1 (b) As shown in Table 7.1 Source: Bukodi, Goldthorpe, Waller and Halpin (2016) largely confined to the working-class positions of Classes 6 and 7 tend to be positively associated with Class 6 and 7 origins (i.e. in Tables 7.3 and 7.4 almost all other classes of origin show significantly negative associations relative to Classes 6 and 7 as the reference category). Further, cognitive ability also proves to be of importance over and above the effect it has on qualifications. For both men and women, class trajectories that are characterised by worklife stability within the managerial and professional salariat are, with one exception, positively associated with being in the two highest ability quintiles, while those characterised by stability within the working class are associated with being in the two lowest quintiles. In other words, the effect of individuals' ability in shaping their class histories would by no means appear to be fully expressed via their educational attainment at labour market entry. In the light of this evidence, extending 146 Social Mobility and Education in Britain over the last half-century or more, one might then conclude that -for better or worse - the realisation of an education-based meritocracy has still some way to go. Such a conclusion is reinforced when we turn to those class trajectories that imply worklife mobility. For men and women alike Type 3 trajectories involve fairly early life upward mobility, including into Classes 1 and 2. However, these trajectories tend, with both genders alike, to be positively associated not with the highest, but with the next-to-highest qualification level at labour market entry (i.e. in Tables 7.3 and 7.4 to be negatively associated with all other levels as compared with this reference category). And it can further be seen that Type 3 trajectories are, again with both genders, positively associated with Class 1 and 2 origins, and for women also with high cognitive ability. What is thus suggested is the possibility that trajectories of this kind may often reflect what has been referred to as 'counter mobility': that is, upward worklife mobility that serves, in the case of individuals from more advantaged class origins, to offset downward intcrgenera-tional mobility at labour market entry, as might be the result of only modest educational performance. This possibility we explore further in Chapter 9.10 Moreover, in two other mobility trajectories, educational attainment proves to have a yet more limited role. Type 5 trajectories for men and Type 6 for women, which involve mainly upward mobility from Class 6 and 7 positions into the small employer or self-employed positions of Class 4, are shown to be positively associated with both the next-to-lowest and the lowest qualification levels -although also with Class 4 origins, and for women with Class 1 and 2 origins. The inheritance of capital or 'going concerns' may thus often play a part in this kind of trajectory but it can also be followed without this advantage as well as without that of a high level of educational attainment.1 10 The concept of counter mobility is due to the Swiss sociologist, Roger Girod (1971). 11 For cross-national evidence of the generally negligible importance of education in regard to access to small employer of self-employed positions of the kind covered by Class 4, see Ishida, Miiller and Ridge (1995). Education and the Labour Market 147 Upward mobility through self-employment Colin Colin grew up in a working-class family. His father always worked shifts at a factory and his mother was an office cleaner. He enjoyed school, especially sports, but left at 16 with few qualifications - which now, looking back, he regrets - but at the time he 'wanted to earn money to buy things', especially a car. His father insisted that he should 'learn a trade' and so he became apprenticed to a roofer. He did not like his boss so left to take up a series of semi-skilled jobs in the construction industry. After some years working 'on the lump' - a form of casualisation of labour through subcontracting - he decided he might as well become independently self-employed. Following a difficult start, during which his marriage broke up under the pressure, he believes, of the financial problems involved when he was unable to obtain a steady stream of work, things steadily got better. He decided to advertise his services locally as a 'roofing contractor' and after that he began to prosper. He now has a well-established business which demands 'continuous hard work' but which has made him 'comfortable' financially. He has a large house and garden and now that his children, who stayed with him after his wife left, have grown up, he lives there alone - 'very contentedly', he says. Likewise, the Type 4 trajectory for women, which similarly involves upward worklife mobility from working-class positions, is positively associated with the two lowest qualification levels and - unlike the trajectories leading to Class 4 - is also associated with the two lowest cognitive ability quintiles. It may then be that in this case mobility is primarily promoted by non-cognitive attributes of some kind, such as personality characteristics. Support for this possibility is found in more detailed analyses we have undertaken which show that seven out of the ten occupations to which the trajectory most frequently leads are supervisory or managerial ones in the personal services or retail sectors. 148 Social Mobility and Education in Brit, Upward mobility with low qualifications Angela Angela enjoyed school, mainly because she excelled at sports: 'I always wanted to win.' She thought of training as a PE teacher but then decided to leave school at 16, with only one or two O levels, 'because I wanted to earn money'. She started work as a post-girl in an insurance company office but while still a teenager married, had a daughter and then divorced. After a period of single motherhood she returned to work in her late twenties in various part-time jobs in retail and services but then applied for, and got, a full-time job as a departmental manager in a furniture store where she had been previously employed and was well thought of. Over several years she 'worked her way up' to higher-level managerial positions in the firm: 'Again, I had to win, to be on top, there's no point, you know, in being down at the bottom.' She was eventually 'poached' by a rival firm for a top managerial position, and bargained a large salary increase: 'I'm good at my job.' She says that, although she still has problems with personal relationships, 'work-wise and money-wise' her life has become 'fantastic'. The only worklife mobility trajectory in which the role of education appears to be quite dominant is in fact the downward mobility trajectory to working-class positions for women, Type 7. Being in the two lowest qualification levels is associated with this trajectory, but there is no association with class origins nor in any clear way with cognitive ability. Overall, then, the ways in which the class trajectories of the members of our three cohorts have been shaped would appear far more complex than envisaged in the liberal scenario of postindustrial society as outlined at the start of this chapter. The one question that remains, however, is that of whether, even if an education-based meritocracy does not as yet exist in Britain, there is evidence of some continuing movement towards it. To address this question, we have repeated the analyses that underlie Tables 7.3 and 7.4 for each cohort separately. What we find may be best described as merely minor variations on the pattern of results we have presented for the cohorts taken togethct. There is no indication of any directional change across the cohorts, and education and the Labour Market 149 certainly not of a kind that could be interpreted as showing an increasing import:ance OI individuals' qualifications at labour market entry for the class trajectories that they subsequently follow.12 In sum, the results we have reported would indicate that, if the cornparison is between cohorts born before and after the Second World War, then entry into higher-level managerial and professional positions has become more strongly associated with the possession of higher-level educational qualifications; but that over the postwar decades further change of this kind, in the direction of an education-based meritocracy, has not been sustained. Or, in other words, there is little reason to believe that education per se - that is, considered independently of class origins and cognitive ability - is any more class destiny today than it was fifty or sixty years ago. In conclusion, it is then relevant to ask why liberal expectations in this regard have not been met, and further what are the implications for current discussion of social mobility in political and policy contexts, which, as we have noted, has been much influenced, even if only implicitly, by theories of postindustrial society that underlie these expectations. To begin with, two misjudgments can be identified. First, the extent and the durability of the effects of individuals' social origins on their class histories have been inadequately appreciated. As we have previously shown, inequalities in individuals' educational attainment in relation to their social origins - that is, in relation to the level and range of their parents' resources - have remained more or less unchanged across our birth cohorts. Now, we have further shown that individuals' social origins have a similarly persisting effect on their chances in the labour market over and above that which is mediated through the educational qualifications they obtain prior to entry. In this way, therefore, the issue of the degree of compatibility between equality of opportunity and inequality of condition again arises. Second, the importance of theoretical knowledge within the post-industrial economy has been exaggerated. Advancing technology and more complex organisational and economic contexts undoubtedly increase the demand for such knowledge in some sectors. But this is far less the case in others - and including in ones in which employment growth has tended to be strong. For example, in Britain since the 1970s employment growth has been most marked not in those sectors that Further details of the analyses undertaken tan be found in Bukodi et al. (2016). 150 Social Mobility and Education in Br, 'tain might be regarded as constituting the 'knowledge economy' but rath in a range of personal services - including various forms of care hospitality, leisure and travel services - and in retail.13 In these latter sectors many jobs do of course involve only quite routine wage work but there are also higher-level supervisory and managerial positions' and these are not for the most part ones that call for any grea| theoretical knowledge as opposed to the kind of practical knowledge that can be readily gained through work experience. Thus, opportunities can still exist for promotion from below for men and women with perhaps few formal qualifications but who can, through their work performance, directly demonstrate their suitability for it. Finally, the yet more basic question can be raised, although it is one that seems little considered in political circles, of the unintended and perhaps unwanted consequences of taking an education-based meritocracy as in effect an ultimate policy objective. It could obviously be thought desirable, on efficiency as well as normative grounds, that selection for more demanding, and better-rewarded, social positions should be based on evidence of relevant competencies. But how far formal educational qualifications are to be seen as providing the only or in all cases the most reliable, such evidence is open to some doubt. The danger always attending the idea of an education-based meritocracy is that of 'credentialism': that is, the danger that certain formal qualifications become a strict requirement for gaining entry to certain occupations, whether or not such knowledge, expertise and skills as they serve to certify are actually required in carrying out the work involved. And what has then to be recognised is that credentialism can in this way constitute a serious and unnecessary barrier to social mobility - reinforcing in the labour market the inequalities of opportunity initially arising within the educational system.14 With this danger in mind, the findings that we have reported of the persistence of trajectories of upward worklife class mobility, with no strong association with level of qualification at labour market entry, Education and the Labour Market 151 Id be viewed quite positively - and could, moreover, point to the L°U rial value of some reorientation of attempts aimed at increasing ability via the practices of employers. So far, the main focus of such has been on employers' recruitment practices, mainly as a attempts "as i c c ■ i i- . I ult of concerns that the use of a range of job applicants personal -haracteristics, as well as their qualifications, as signals of their suitability is likely to 'eac* to soc>a' biases that impede mobility. But a focus on employers' promotion practices might prove to be at least as, it not more, rewarding: that is, with the aim of discouraging credentialism that effectively blocks promotion from below for those without some, perhaps quite arbitrarily determined, level of qualification, and of encouraging the wider development of internal promotion programmes and associated training provision.15 In this way, greater opportunity could be created for upward mobility over the course of working life on the part of men and women who have actually shown themselves to be capable of moving on to a higher grade of work, and regardless of their possession or otherwise of'passports of recognition' in paper form. The Social Mobility Employer Index, sponsored jointly by the Social Mobility Commission and the Social Mobility Foundation, aims to rate employers - but only those who volunteer to be evaluated - according to their procedures in regard to both recruitment and 'progression'. The emphasis so far has been heavily on recruitment, although the Commission has recently shown some greater concern over the improvement of 'internal pathways to promotion' and especially for men and women with low level qualifications (Social Mobility Commission 2016: 147-54). The particular occupations in which numbers increased most in the last quarter of the twentieth century include care assistants and attendants, educational assistants and hospital ward assistants - all in the top ten - with travel and flight attendants, hotel porters, merchandisers and window dressers not far behind (Goos and Manning, 2007). Concerns over credentialism have been far more marked in the US than in Britain, stemming from the influential work of Collins (1979).