Grusky - Part IV Generating Stratification

Social Mobility

Classical Viewpoints

Pitirim A. Sorokin - Social and Cultural Mobility

This article is pretty straightforward. Sorokin defines social mobility as any transition of an individual, social object or value from one social position to another. There are two principle types of mobility: horizontal and vertical. Horizontal is movement between positions on the same level (ex. from Methodist to Baptist, same job in different companies, etc.). Vertical social mobility is the movement between one social stratum to another. This is what we are most concerned with.

There are two types of vertical mobility: ascending (or social climbing) and descending (or social sinking). Two forms of ascending movement: the infiltration of people in lower stratum into an existing higher one; and the creation of a new group by the upwardly moving persons (may be instead of or side by side with the existing higher stratum). Two forms of descending movement: the dropping of individuals from a higher position to a lower one; and the dropping of a whole social group.

There are two theoretical extremes of stratified societies. One with zero vertical stratification. This is a completely immobile, rigid system. The opposite extreme is a system with intensive and general vertical mobility. This Sorokin calls a completely mobile or open society. Between these two extremes there are many intermediary types.

Sorokin lists several general principles of vertical mobility:

1. There seems to be no definite perpetual trend in any society toward either increasing or decreasing mobility. What happens is that the obstacles that prevent vertical mobility in previous times such as the feudal system, may be abolished, but they are simply replaced with a different type of impediment.

2. "In present Western societies different occupational groups are strongly interwoven, and the cleavages between them are considerably obliterated, or more accurately, are somewhat indefinite and not clear cut" (pg. 247). There is a fluid composition to modern occupations, meaning that one is not limited to the occupation of one's father, for ex. This dispersion of individuals across occupations reduces the separation between these occupational groups.

3. However, despite the dispersion described above, 'hereditary transmission' still exists and is fairly common, on average. Children are still likely to enter their father's profession in higher proportions.

4. "The closer the affinity between occupations, the more intensive among them is mutual interchange of their members; and, vice versa . . ." (Pg. 247). Sorokin argues that this means that cleavages are more likely between larger social subdivisions of occupations than between occupational groups in a narrow sense.

Sorokin goes on to describe the solidarity of societies that are immobile vs. those that are mobile. In immobile societies, social solidarity of members is concentrated in each of the social "boxes" in which they live because social contact with people in other boxes is weak and rare. So there is little or no solidarity between boxes, but with boxes the ties are very extensive. In a mobile society, however, individuals belongs to different groups and move between boxes. Each members area of social solidarity is not limited to a single box. Sorokin says this atomizes and individualizes solidarity because each persons ties are to different people in a variety of boxes. In addition, he says that these conditions reduce the intensity of the attachments. This has two consequences: 1) the map of solidarity and antagonisms are more complex in a mobile society than in an immobile one; 2) these lines of solidarity are also more flexible and changeable.

Finally, Sorokin argues that mobility favors increased individualism. It separates the individual from the social box. "This participation in many groups, shifting from one group to another, and impossibility of identification with any one group makes an individual something separate from a social box; awakens his personality, transforms him from the component of a group to an individual person" (pg. 249). However, this individualism is accompanied by a search for attachment to some social body in order to avoid loneliness and isolation. Such attachments include churches, social clubs, political parties, labor unions, dance halls, etc. These "collectivist tendencies" are the flip side of the individualism created by mobility. They are attempts to find substitutes for the closely-knit social boxes of immobile society that have been lost.

Lipset, Bendix and Zetterberg - "Social Mobility in Industrial Society"

This article has a more empirical focus. The authors attempt to summarize the findings of mobility research for a number of countries (U.S., Germany, Sweden, Japan, France, Switzerland, Denmark, Great Britain, and Italy) and make a few conclusions from these findings. The begin with the observation that widespread social mobility appears to be a basic characteristic of modern industrial society. In fact, they find that the patterns of mobility in various Western countries appears to be the same. In addition, they find little support for the idea that rates of mobility and rates of economic expansion are correlated. Instead, their interpretation of the findings "is that the social mobility of societies becomes relatively high once their industrialization, and hence their economic expansion, reaches a certain level" (pg. 251).

Data collected since WWII allows the authors to make comparisons between countries of variations in occupational mobility. They acknowledge that making such comparisons is difficult and problematic. To simplify things, they use a three category occupation distinction (manual, nonmanual, and farm occupations) and they focus only on males. Initial examination shows that total mobility rates in these Western countries is virtually identical. However, there appears to be significant variation between countries in the degree to which father's occupation is an asset or handicap.

The authors list several processes they say are inherent in all modern social structures and have a direct effect on rate of social mobility. These processes help account for the striking similarity in mobility rates across countries.

1. Changes in the number of available occupational vacancies. Industrialization increases the proportion of workers who are professional, managerial, white-collar, etc. This along with a decrease in the proportion of unskilled labor jobs crease a surge of upward mobility.

2. Rates of fertility. In industrialized countries, fertility tends to be negatively associated with income. So, if every high status person's son retains that status, there would still be room for others to move up.

3. Changes in the rank of occupations. In a rapidly changing world, some occupations gain prestige and others lose it. This means that a person can be mobile (move up or down) without changing jobs at all.

4. Changes in the number of inheritable positions. Modern social structures usually have fewer inheritable positions. Since people are no longer guaranteed these positions by inheritance, there is room for others to move in.

5. Changes in legal restrictions on opportunity. The abolition of the guild system is an example of this. Many earlier legal restrictions on the right of a person to create a newer and higher occupational status have been removed.

Next the authors address the consequences of social mobility. The argue that although the amount of social mobility is similar across countries, the consequences of that mobility can quite varied. The key concept needed to understand these variations is status discrepancies. Each society has a number of hierarchies (social, educational, ethnic, etc.). There are likely to be a number of discrepancies in status among the positions in these hierarchies that a person holds simultaneously. "Mobility merely adds to these discrepancies by creating or accentuating combinations of a high position in one rank and a low one in another" (pg. 255). An example would be a highly educated Black woman. Status discrepancies can cause difficulties in personal adjustment. Mobility also has disruptive consequences on primary group structures from which individuals gain support through family, close friends, etc. Across the countries examined, the authors note that in Northern European countries where changing classes involves a major adjustment in lifestyle, upwardly mobile individuals are more likely to maintain links with their class of origin. In the U.S., though, presumably upwardly mobile individuals find it easier to adjust.

This leads us to the authors final point. They argue that the data they present raises questions about the commonly accepted belief that the U.S. is the land of opportunity. They say that the absence of a feudal past has not meant that the U.S. does not recognize status distinctions like European countries. Rather, it has lead to an ideological equalitarianism despite the numerous and invidious status differences that in reality exist. "The point to emphasize is, not that these beliefs are often contradicted by the experience of those who hold them, but that this equalitarian ideology has persisted in the face of facts which contradict it" (pg. 257). The authors note that the interpretation of ideological equalitarianism varies considerably. One interpretation is that it is a delusion which we must dispel with hard facts (ex. Warner). Myrdal is more optimistic. He argues that the deep moral conflict white Americans experienced as a result of racial segregation can be the source of progressive social change. The authors find both interpretations unsatisfactory. They propose that ideological equalitarianism is not just a belief but part of reality. They argue that Americans think of differences in status and power in terms of differences in the distribution of material goods. This materialistic ideology is unlike class and status ideologies in Europe because contains an idealistic belief in equality. The authors say that Americans will hold on to this ideal in the face of great difficulty (like the Great Depression) and only a sharp and lasting decline in opportunity would threat these beliefs and lead to a change in the value system. We have not yet experienced such a decline.

Ralph H. Turner - "Sponsored and Contest Mobility and the School System"

This paper suggests a framework for examining the relationship between school systems and the norms of upward mobility, comparing the American and English systems. Turner argues that "the accepted mode of upward mobility shapes the school system directly and indirectly through is effects on the values which implement social control" (pg. 260). The norms of upward mobility are presented as two ideal types: contest mobility (CM)is a system in which elite status is determined in an open competition (this is the American form); sponsored mobility (SM) is a system in which elite recruits are chosen by the existing elite (this is the British form). Elite status is given by the established and cannot be taken as it is in CM. (It's like getting into a private club, a current member must sponsor you).

Every society must deal with the problem of maintaining loyalty to the social system. They do this partly through norms and values. There are social control methods that fit CM systems but different ones that work with a SM system. Take the significant problem of ensuring the loyalty of the disadvantaged classes. In a CM system, this is done by combining a futuristic orientation, a norm of ambition, and a sense of fellowship with the elite. Everyone is encouraged to see themselves as one of the competitors. Under such social control mechanisms, Turner notes that organized deviancy is more likely to take the form of an attack upon the conventional moral order rather than an attack on the class system itself. There are very few active revolutionaries.

However, this social control system would fail miserably in a SM system. In such a system, the primary threat would be the existence of a strong group that attempts to take the elite positions. To control this problem, the masses are trained to see themselves as relatively incompetent to manage society. Access to the skills and manners of the elite are limited and a belief in the superior competence of the elite is instilled. The selected elite recruits are thoroughly indoctrinated into elite culture and kept under close supervision. The SC system provides a good setting for developing a sense of "responsibility" for "inferiors" and for the preservation and appreciation of the "finer things of life". In a CM system, elite control is more difficult since there is no controlled system of induction and apprenticeship. The insecurity of the elite position seems to be the principle form of regulation. Success is relative and not guaranteed and competition continues at higher and higher levels.

Regarding formal education, Turner argues that there are constant strains to shape the educational system to fit with whichever norm of mobility prevails (remember these are ideal types, no real society fits perfectly). The strains operate directly, by blinding people to alternatives and influencing their opinions on the resolution of recurring educational problems. They also operate indirectly, "through the functional interrelationships between school systems and the class structure, systems of social control, and other features of the social structure. . ." (Pg. 263).

Turner argues that the different policies for student selection in American and English schools is the most obvious application of the distinction between the influence of CM and SM. In American schools, we have been very concerned with avoiding any sharp social distinction between superior and inferior students. To do so would offend the sense of fairness that accompanies the contest norm. Schooling is seen as an opportunity and making use of it depends on the student's initiative and enterprise. On the other hand, the English system sorts promising students out early in their educational career so they can be segregated and given special training that will prepare them for a higher social standing. This practice fits perfectly with the logic of sponsorship.

Finally, Turner discusses the effects of mobility on personality. He notes that three aspects of the 'mobile personality' are generally stressed: 1) the stress involved in seeking high status; 2) the difficulty in having to abandon lower status friends and gain acceptance in higher status circles; 3) figuring out a personal scheme of values when one moves between two classes that probably have very different and/or contradictory value systems. Turner argues that these three mobility problems have different impacts depending on whether the system is CM or SM. In SM, selection is made early so there should be relatively little strain. Turner concludes that "the factors stressed as affecting personality formation among the upwardly mobile probably are rather specific to the contest system, or to incompletely functioning sponsorship system"( pg. 264).

Modern Analyses

Featherman and Hauser - "A Refined Model of Occupational Mobility"

In this paper, F&H describe their loglinear model of the mobility table. This model focuses on the net or underlying patterns of association rather than gross flows of manpower(which they say class theorists focus on). They are talking about relative mobility. "The model permits us to locate groups or clusters of cells in the classification that share similar chances of mobility or immobility, freed of the confounding influences of the relative numbers of men in each origin or destination category and of changes in those relative numbers between origin and destination distributions" (pg. 265). They point out that they have no theoretical expectations, but their findings have led to substantially different conclusions about the U.S. mobility process than the investigation done by Blau and Duncan.

To clarify, the mobility table refers to a cross-tab of father's occupation and son's occupation(first full-time position). F&H distinguish between five occupational categories: upper nonmanual, lower nonmanual, upper manual, lower manual, and farm. So, the diagonal would consist of people who have the same occupational category as their father, or occupational immobility. The authors refer to the model as "a refined multiplicative model of the mobility table" and I do not think greater detail is important. Therefore, I will skip to the conclusions.

F&H present their model as a cogent and parsimonious alternative to the Blau & Duncan interpretation. The benefit of their model is that "it does not reflect the shape of occupational distributions of origin or destination, but only the underlying patterns of immobility and exchange between occupational strata" (pg. 272). They review the major findings:

1. There is great immobility at the top (upper nonmanual) and at the bottom (farm) of the occupational hierarchy. This immobility seems to be extreme than previously supposed.

2. Surrounding the extremes of the hierarchy are transitional zones within which there are relatively homogenous chances of immobility and exchange (mobility) with the extreme strata.

3. As aggregates, the extreme and transitional zones are relatively closed to upward and downward movements. This means that there appear to be barriers to movement across class boundaries.

4. One the boundary into the transitional zone has been crossed, no social distance gradient appears to underly variation in long-distance mobility chances.

5. While immobility is great at the extremes of the hierarchy, it is almost completely absent in the middle of the occupational hierarchy. Upper blue-collar workers seem to have a lot of opportunity for movement and there do not appear to be "class" boundaries limiting movement to or from skilled manual occupations.

6. There is a rough equality in the movement in one direction with movement in the other. In other words, the mobility pattern is symmetric, no greater tendency towards movement in one direction over the other.

7. Lastly, they note the simplicity of their model.

Finally, they note that these description of relative mobility does not conflict with their previous description of occupational inflow/outflow patterns. The inflow/outflow patterns are affected by the distributions of the origin and destination categories, an effect which has been controlled in the relative mobility model. "Major features of the inflow and outflow tables, like the pervasiveness of upward mobility between generations and within the occupational career, are absent from the present account because they are functions of changing distributions of occupational origins and destinations" (pg. 273).

Grusky and Hauser - "Comparative Social Mobility Revisited: . . ."

This article starts with the Lipset and Zetterberg observation discussed above, that mobility rates are much the same in all Western industrialized societies. G&H note that recent data does not support this position. They then describe the FJH (Featherman/Jones/Hauser) revision, that observed mobility rates may derive from historical and cultural variation, but not from differences in exchanges between occupations. The FJH model predicts that mobility chances are invariant once variations in origin and destination distributions are controlled (see above article). However, up to this point cross-national comparisons with large samples support variability. Another group of mobility scholars have long contended that mobility increases with industrialization, even after class or occupational distributions are controlled. This contention is referred to as the thesis of industrialization and is generally set up in opposition to the FJH revision. The FJH allows for an initial developmental effect on mobility, but implies no effect once a certain level of industrialization is reached. G&H point out that evidence has been inconclusive regarding both theses.

The purpose of this paper is to address both theses and to explore the role of noneconomic variables in mobility processes. They use a three-stratum classification (white-collar, blue-collar, farm) to create 3x3 mobility tables (father's occupation by son's occupation), which they argue is sufficient to capture important barriers to occupational mobility (and other differences in life chances).

They discuss several equivalent 3x3 models that they see as having a satisfactory fit and are theoretically interesting. One is the quasi-perfect mobility model which specifies association between origins and destinations in terms of parameters for inheritance. G&H use this model to examine differences between occupational strata in opportunities for mobility or inheritance, which they believe are based on variation in the resources and desirability of an occupation. From this model we would expect strong white-collar immobility and considerable mobility for son's of blue-collar workers. Farm immobility is expected to be even strong than that of the white-collar stratum. There are two implications that follow: 1) relative strength of stratum inheritance may be uniform across countries simply because there is substantial uniformity in resources and desirability of occupations; rather than coming from industrialization, the common structure of mobility may apply to all societies regardless of economic development.

Now, to the analysis. They use a 16 country sample to make cross national comparisons of mobility in order to address the above theses. First, they reject the Lipset-Zetterberg hypothesis (which apparently has never been directly tested before). They find that "not only is there highly significant variation in observed mobility rates among industrialized nations, but there is no less variation among these nations than among nations that vary widely in level of industrialization" (pg. 277).

Next, they address the FJH revision. Their analysis shows considerable support for the FJH revision. In fact, the invariance (controlling for historical and cultural variation) appears to be even stronger than expected. In addition G&H propose two elaborations on the FJH revision. First, they suggest that the uniformity in mobility is not limited to industrialized countries, but is common in all societies regardless of level of economic development. Second, they add to FJH by specifying the structure of the common mobility regime: the quasi-perfect model. They find a symmetry of exchanges between strata once intergenerational shifts in marginal distributions are controlled. In addition, they find a similar picture in all 16 countries examined: severe immobility at the two extremes of the occupational hierarchy and considerable fluidity in the middle.

Erikson and Goldthorpe - "Trends in Class Mobility: The Post-War European Experience"

This article addresses the liberal theory of industrialism which became the dominant view in the 1960s and was developed by various American authors (Kerr, Dunlop, even our buddy Parsons). This theory can be laid out in a 3-part proposition:

In industrial societies (in comparison to pre-industrial societies):

1.Absolute rates of mobility are generally high and upward mobility predominates over downward mobility.

2. Relative rates of mobility, or mobility opportunity, are more equal. So, individuals of different social origins compete more equally.

3. Both the level of absolute rates of mobility and the relative rates of mobility (degree of equality) increase over time.

E&G then go over several explanations for why these three changes between pre-industrial and industrial societies should occur. They see all of these arguments as functionalist, but distinguish them in terms of three types of effects: structural, processual, and compositional.

Structural arguments hold that the dynamism of rationally developed technology calls for continuous, often rapid, change in the social division of labor which also becomes increasingly differentiated. High rates of mobility stem from this redistribution. The increase in bureaucratic and administrative apparatus creates the tendency toward upward mobility. In industrial society, the chance for success steadily improves.

Processual arguments claim that industrialism transforms the processes by through which people are allocated to different positions. This involves a shift away from ascription towards achievement (see Magnew's critique of this).

Compositional arguments say that structural and processual effects interact with one another. The emphasis on achievement is strongest in the expanding sectors of the economy and ascriptive procedures will persist in declining sectors and organizational forms. So, the proportion of the population subject to the new mobility regime characteristic of industrialism steadily increases further as those areas that are most resistant steadily become more and more marginal.

E&G attribute the liberal theory's dominance to the coherence of its development and the great failure of the main attempt to controvert it: the revision and extension of the Marxist theory of proletarianism. This approach lacked from the start any secure empirical support and finally fell apart with the accumulation of results incompatible with it. The authors do point out several other positions that still can be seen to conflict to some degree with the liberal theory that are worth discussing.

1. Lipset and Zetterberg - we should know this by now. Although it is often assimilated into liberal theory, but there are several crucial differences. Most importantly, L&Z do not argue for steadily increasing mobility with industrialization. Instead they propose a threshold effect. Mobility becomes relatively high once a certain level of development is reached. Although their contention of uniformly high mobility would not now be supported, "their suggestion that a historic upward shift in such rates tends to occur at some-perhaps quite early-stage in the industrialization process has not been similarly disconfirmed" (pg. 291). L&Z do not suggest industrial societies are more open (more equal opportunity), they firmly emphasize structural change.

2. Sorokin (see earlier article) - Sorokin was against the idea of a perpetual and eternal increase in mobility. Rather, he saw the increased mobility modern societies were experiencing as simply an historical phase. There is only an endless alternation between greater and lesser mobility. One of the most useful points we gain from S.'s position is the idea that while some barriers to mobility have been removed, others have become more severe or new ones added (ex. systems of educational selection and occupational qualification). No matter how the system works, those in power will do everything they can to maintain their privileged position. This is tendency for closure is endogenous to all forms of stratification. Sorokin also points out that there are exogenous factors as well (political and economic upheaval associated with war or technological change, for ex.). In these sorts of upheavals, an surge in mobility is produced as the entire system is disrupted.

3. FJH - here they are again. Theirs is a revision of the L&Z thesis: across industrial societies rates of social mobility are basically the same. E&G argue that this hypothesis does not stand if applied to absolute rates, but is much more plausible in the case of relative rates (opportunity). E&G find the FJH hypothesis useful as a comprehensive challenge to liberal theory. "It implies a basic skepticism, essentially akin to that of Sorokin, about the possibility of any long-term, developmentally-driven trend; while, on the other hand, it stands directly opposed to the proposition that under industrialism a steady increase occurs in the equality of mobility chances" (pg. 292).

Okay, I've gone to such detail above because I think it is a good review of the approaches to mobility. Now, we get to the purpose of the article at hand. E&G argue that there is a certain American ethnocentricity within the liberal theory. Most studies make either an implicit or explicit US/Third World comparison. So, the industrialization process of mainland European nations is examined in a very restricted manner. The supposition is that, since England industrialized early, the rest of the nations followed along behind on the same path, differentiated only by the amount of time lag. E&G argue that this assumption has no sound historiographical basis. Rather, there has been a great diversity in developmental paths and we should remember that later industrialization does not automatically mean economic retardation. Also, it is a mistake to assume that Europe was completely industrialized by the end of the 19th century. In fact, industrialization was not complete until after WWII. What E&G intend to do is use the rich data on Europe to make inferences about mobility trends in general.

Pages 296 to 303 provide several tables showing the analysis of absolute mobility rates for 9 European nations. They also use three strata: farm, manual, and non-manual, but also specify service class positions. .E&G summarize their findings regarding absolute rates. These are the trends that appear likely to have occurred in all or most nations at some stage in their development:

1. A decline in intergenerational mobility within the farm sector.

2. An increase in mobility from farm origins into manual industrial employment.

3. Some increase in mobility from farm and manual origins into service-class positions.

They conclude that there appears to be some increase in broad mobility flow associated with industrialization, but they have found these trends to be rarely continuous over time and the timing varied between nations. Also, changes in all other rates examined appear to be directionless. So overall, there is little evidence that mobility rates are moving in some "industrial pattern" the way liberal theory predicts and there is certainly no trend towards ever increasing mobility.

Regarding relative rates, E&G use a model based on predictions of the FJH hypothesis that they call the constant social fluidity (CnSF) model. They find that this model fits fairly well.

So, what do we conclude? E&G argue that considerable doubt as been place on the liberal theory of industrialization by their analysis. There is no evidence of general and consistent trends toward higher levels of either total mobility or social fluidity. No evidence that either absolute or relative mobility rates change in any consistent direction or that these rates become more similar between nations over time. The most that can be said: structural changes, particularly the decline in agriculture, are likely to create an upturn in total mobility for a limited time and at various phases of development. Instead, what they find is: absolute mobility rates are characterized by trendlessness, while relative mobility rates are characterized by considerable stability. The trendlessness of absolute rates raises the question of whether they are even a phenomenon open to macrosocial explanation. Regarding the constancy of the relative rates, E&G note: "While we have not been able to support the claim of a sustained developmental trend, we have, it appears, found indications of something of no less sociological interest: that is, of a constancy in social process prevailing within our several nations over decades that would in general have to be characterized in terms of the transformation and turbulence that they witnessed" (pg. 310). E&G connect this finding of constancy with the FJH hypothesis. The constancy they found prevents them from ruling out the FJH prediction of commonality.

Status and Income Attainment

Basic Models

Blau and Duncan - "The Process of Stratification"

This article addresses the issue of ascription vs. achievement in modern society. Generally, both ascription and achievement principles are recognized as existing, but we tend to think of achievement as more powerful and a more basic aspect of our society. Ascriptive features are regarded as left-overs from previous eras that will diminish over time. Public policy focuses on ways to improve or equalize opportunity. But how far can society actually move in the direction of achievement? This question is hotly debated. B&D's contribution, they argue, is in "submitting measurements and estimates of the strength of ascriptive forces and of the scope of opportunities in a large contemporary society. The problem of the relative importance of the two principles in a given system is ultimately a quantitative one" (pg. 317).

To do this, B&D conceptualize an individual's life cycle as a sequence that can be described, to some extent, by quantitative measurements or classifications taken at successive stages. The questions they are asking boil down to: To what degree do circumstances of birth condition subsequent status? How does status attained either, by ascription or achievement, in one stage of the life cycle affect later stages? By selecting various quantitative variables that they say are sufficient to describe the major status changes in the life cycle of a cohort, B&D argue that they have built a basic model of the process of stratification. The variables they select are: father's educational attainment, father's occupational status, respondent's educational attainment, status of respondent's first job, and status of respondent's occupation in 1962. The occupational status measurements are based on B&D famous socioeconomic index (which is discussed in Part III). They use path coefficients to examine the effect of each variable on the others, direct and indirect. I've inserted a copy of Figure 1, pg. 321, which shows the basic model.

They give an excellent, straightforward discussion of path coefficients on pp. 318-325, which I will skip here. I will only make one note: the form of the model and especially the numerical estimates made are valid only for the population under study (in this case the U.S. population). Path analysis is not a method for discovering causal laws. Rather, it provides an interpretation for a known or assumed causal system in a particular population. Only by comparison can we know if the same interpretative structure works for more than one population (i.e. the populations of other nations or even for a subpopulation of the U.S.).

So what have we discovered from this investigation. B&D discuss the idea of a vicious circle which they see popularized by the ideas of a "cycle of poverty". This issue is laden with ideological beliefs and characterized by rhetoric. B&D begin with the question of evaluating the facts. In other words, how 'difficult' is it for a person to change their status from the status of their family of orientation? They find evidence for a considerable amount of status modification, or occupational mobility. How much modification must occur to contradict the vicious cycle diagnosis is a judgment call.

Regarding the influence of family status (measured by father's ed and father's occup), B&D find that family status is associated with education which in turn affects occupational achievement. But, of the effect of education on occupational status, only a small proportion comes from the indirect effect of family status. Most comes from the direct effect of education itself. In fact, education serves to induce variation in occupational status that is independent of family status.

To summarize the quantitative findings, here is a breakdown of the percentages of variation in respondent's 1962 occupational status attributed to each set of variables:

(i) Total effect of father's education and occupation 18.06

(ii) Effect of respondent's education independent of (i) 24.32

(iii) Effect of all other factors, independent of (i) and (ii) 57.62

Total 100.00

B&D's conclusion: They find the influence of family background (or ascription) to be less determining than that of achievement. "The relationship is not trivial, nor is it, on the other hand, great enough in itself to justify the conception of a system that insures the 'inheritance f poverty' or otherwise renders wholly ineffectual the operation of institutions supposedly based on universalistic principles" (pg. 327). They point out that vicious circle arguments say that the intercorrelation of background variables compound the effect. Their analysis shows, however, that these intercorrelations result primarily in redundancy, not cumulation. They note that they do not deny the existence of vicious circles altogether. They see the situation of blacks in America as an example of how these circles do actually occur. What they want to do is point out that vicious circles are distinct phenomena that should not be confused with the basic model of status attainment. Finally, they argue that they are not simply supporting a functionalist interpretation that says that American society has just the right amount of stratification necessary and just the right degree of intergenerational status transmission (See, for ex., Kingsley and Moore in Part II). Rather, they have indicated that it is easy to exaggerate the effect of family status and misrepresent the nature of the causal relationships in the status attainment process.

Jencks et al - "Inequality: A Reassessment of the Effect of Family and Schooling in America"

The author begin by noting that most Americans say they believe in equality, but their definitions are usually full of contradictions. They say that all people are equal, but they generally also believe that some people are more competent than others and will always be no matter how we reform society. In addition, competency is expected to be rewarded by success, while incompetence should be punished with failure. There is no commitment to real equality (like equality in political power, in income, or in desirability of one's job). But while accepting this inequality in everyday life, Americans still believe in "equal opportunity". Equal opportunity meaning that the rules of the game, that decides who succeeds and who fails, should be fair.

It is this idea of equal competition that was addressed by most reformers in the 1960s. Only a very few radicals talked about eliminating inequality itself. There was no attempt to reduce disparity directly. The authors argue that the 1960s reforms failed precisely because they ignored the problem of inequality directly. "Equalizing opportunity is almost impossible without greatly reducing the absolute level of inequality, and the same is true of eliminating deprivation" (pg. 330).

A main difficulty in these discussions is the definition of poverty. Most official publications define poverty as an absolute condition, rather than a relative one. Most Americans, however, define poverty in relative terms. Public opinion surveys tell us that when people are asked how much money an American family needs "to get by", they typically give an amount that is roughly half of what the average American family actually receives. Political definitions of poverty tend to follow these popular measures and have consequently changed over time. As the cost of living increases, so does the poverty level. What the authors derive from this is that the cost of living is not the cost of buying a certain number of fixed goods and services. Rather, it is the cost of participating in the social system. "The cost of participation depends in large part on how much other people habitually spend to participate. Those who fall far below the norm, whatever it may be, are excluded. It follows that raising the incomes of the poor will not eliminate poverty if the incomes of other Americans rise even faster" (pg. 331). It is the distribution that is important, not the absolute amount. The authors argue that this logic applies to the skills gained from schooling as well.

So, if we think of poverty and ignorance as relative rather than absolute conditions, then eliminating or reducing them means eliminating or greatly reducing inequality itself. They give two options for reducing inequality: make the system less competitive by reducing the rewards for success and the costs of failure; or we can make sure that everyone enters the competition with equal advantages and disadvantages.

The war on poverty in the 1960s focused on education, trying to give everyone comparable skills. People thought that equalizing cognitive skills would equalize bargaining power resulting in a system in which no one was very poor and no one very rich. This strategy rested on several assumptions: 1) eliminating poverty is mostly a matter of helping children born into it to get out of it; people did not fall back, and middle class children rarely ended up poor; 2) the primary reason poor children remained poor was a lack of cognitive skills; 3) educational reform is the best mechanism for breaking this vicious circle.

The authors find each of these assumptions to be erroneous: 1) poverty is not primarily hereditary, inequality is recreated anew in each generation; 2) lack of cognitive skills is not the primary reason people are poor, there is as much variation economic situation among those who test high on standardized tests as there is in those who score low; 3) there is no evidence that school reform can significantly reduce cognitive inequality.

Therefore, many popular explanations of economic inequality are just plain wrong. The authors find that economic success appears to depend primarily on luck and on the job competence that are only moderately related to family background and cognitive skills. So, we must address the problem of inequality in a different way. Rather trying to change people, we need to change the rules of the game in order to reduce the rewards of success and the costs of failure. We need to create "insurance" systems that neutralize the effects of luck, and income-sharing systems that break the link between vocational success and living standards. They suggest several ways this could be done: limiting the disparities in wages that employers pay the best and worst paid workers, income supplements, free public services. Basically, the lucky and the competent would subsidize those with the least competence and luck. This strategy was rejected in the 1960s because it lacked popular support, but that does not mean it was the wrong strategy. "It simply means that until we change the political and moral premises on which most Americans now operate, poverty and inequality of opportunity will persist at pretty much their present level" (pg. 333). The authors begin with the premise that everyone's happiness is of equal value. Therefore, society should be organized to provide the greatest good for the greatest number. They do not suggest a complete equalization which would lead to a lack of motivation. They see a great need for incentive and they prefer monetary incentives over social or moral ones. But it is not necessary to have as extreme a degree of inequality as we now have. For example, if we assume that the most productive fifth of workers accounts for half the GNP, it does not mean they need to receive half the income. A third or quarter would probably be enough to keep them comfortable, happy and productive while allowing the rest of the population a greater degree of comfort and happiness as well. The economic distribution can never be made completely equal because of differences in cognitive and vocational competence. We can and should, however, be working to reduce the inequality significantly.

Social Psychological Models

Sewell, Haller, and Portes - "The Educational and Early Occupational Attainment Process"

This article builds on the work of Blau & Duncan discussed above on the status attainment process of American adult males. What the authors find unsatisfying about this model is the exclusion of social psychological (SP) variables. The B&D model includes only father's education and occupational status, respondent's education and the dependent variable is respondent's occupational status. They argue that although it remains to be seen whether such SP variables have an impact, it seems likely that they do, particularly given the interpretation of the associations between variables that B&D provide. SH&P argue that B&D's model, by omitting SP variables, "fails to indicate why any connection at all would be expected between the input variables, father's education and occupation, and the three subsequent factors: respondent's education, respondent's first job, and respondent's 1962 occupation" (pg. 337). The authors' intention in this article is to remedy this situation and incorporate SP variables into the B&D model. The diagram of their model on pg. 338 is attached, which shows the hypothesized relationships.

The SP variables that they add are: mental ability (measured in high school); academic performance (in high school); significant others' influence, which may be parents, friends, etc. (also measured in high school); level of educational aspiration; and level of occupational aspiration (as a youth); plus they leave room for the effect of exogenous or unmeasured variables. Like B&D they perform a path analysis in which they calculate every possible path in their model (accounting for causal or temporal order), but predict that 8 or 10 paths, shown in the diagram, should be considerably more significant than the rest. The hypotheses are: 1) mental ability affects academic performance; 2) family's socioeconomic status affects significant others' influence and possibly academic performance; 3) academic performance affects significant others' influence; 4)significant others' influence affect level of occupational aspiration and level of academic aspiration; 5) educational aspirations influence educational attainment; 6) occupational aspirations influence occupational attainment; and educational attainment influences occupational attainment. In addition, various unmeasured exogenous factors are expected to affect academic performance, significant others' influence, occupational aspiration, educational aspiration, educational attainment, and occupational attainment. These are direct effects. Indirect effects are located by following the paths between the variables. See the diagram.

The above may be a little difficult to follow. Hopefully the diagram clarifies. Just remember that what SH&P are doing is inserting social psychological variables into Blau & Duncan's model to explain how the process occurs. B&D just say that, for example, family background effects education which in turn effects occupational status. SH&P refine this by showing the mechanisms through which the effect occurs (like significant others' influence and aspirations) as well as explicitly including mental ability which was previously discussed but not tested.

The path analysis shows strong support for the social psychological explanation. In particular, significant others' influence had significant effects on the 3 following variables (educational aspiration, occupational aspiration, and educational attainment) which ultimately influence occupational attainment. None of the unpredicted paths turned out to be very strong. The author's make several closing comments: 1) the significant others' influence variable is clearly an important factor and it is also amenable to experimental testing; 2) Aspirations appear to perform mediating functions, but experimentation is need here; 3) to what extent is this system inherently culture-bound? Would the process occur similarly in Turner's "sponsored" rather than "contest" achievement system? The data cannot answer these questions, but they are worthy of discussion. The authors think that "by allowing for variations in the path coefficients, the same basic social psychological model might work well to describe attainment in stratification and mobility systems quite different from that of the present sample; 4) the linear model used appears to be appropriate; 5) however, despite the satisfactory performance of the linear model, there is still the possibility that techniques based on other assumptions might perform better; 6) we should remember that the most highly educated respondents had just begun their careers (due to the time spent in school) and if the distance between them and the less educated widens, the amount of variation explained by the model may increase; 7) this model was done on a sample of boys from farm backgrounds. It needs to be repeated on populations with more differentiated backgrounds.

Jay McLeod - "Ain't No Makin' It: Leveled Aspirations in a Low-income Neighborhood"

This is a nice little ethnographic piece that I really enjoyed. It was a nice change from the quantitative approaches that we have been inundated with in this section. McLeod studied teenage boys from a housing project. He describes two groups: the Hallway Hangers and the Brothers. Before a detailed description, McLeod makes an important conceptual distinction between aspirations and expectations. "Aspirations are one's preferences relatively unsullied by anticipated constraints; expectations take these constraints squarely into account" (pg. 348).

The Hallway Hangers - This group consisted mostly of white boys. Their name derives from there most common activity: hanging around in the hallways of the projects. These boys have very little aspirations or expectations for economic success. Work is a peripheral issue for them. It is simply a means to earn money, it is not important in itself. In describing what they hope or expect for their futures, work is rarely even mention. "Their expectations are not merely tempered by perceptions of the opportunity structure; even their aspirations are crushed by their estimation of the job market" (pg. 348). McLeod notes that it is difficult to tell if the Hallway Hangers' leveled aspirations are the result of negative assessments of their own capabilities or of the opportunity structure. Both probably play a role, but McLeod sees the strong negative view of the opportunity structure as playing the most important role. In all the levels of experiences they have with the job market (community, family, and personal), these boys are taught that the job market does not necessarily reward talent or effort.

The Brothers - this group is almost completely made up of black boys. They have a very different outlook from the Hallway Hangers. The Brothers are much more positive. They are hoping for a brighter future and work is a central concern. All of their hopes for the future involve good jobs. Unlike the Hangers, the Brothers do not feel that nurturing aspirations is futile. They also view lack of success on the job market as a reflection of personal inadequacy rather than lack of opportunity. "These viewpoints are consistent with the dominant ideology in America. . . By attributing failure to personal inadequacy, the Brothers exonerate the opportunity structure. Indeed, it is amazing how often they affirm the openness of American society" (pg. 354).

So what do these observations lead us to? McLeod argues that the findings are a major challenge to economically determinist theories. These two groups of boys are economically identical. So how do we account for the extreme variation? If social class is "the overriding determinant in social reproduction, what accounts for the variance in the process between the Brothers and Hallway Hangers?" (pg. 355). McLeod suggests that Bourdieu's notion of habitus can help us understand these differences. He argues that habitus, as an explanatory tool, is quite useful here because it is flexible enough to accommodate the interactions between ethnicity, family, schooling, work experience, and peer associations that he has described. However, he does not agree with Bourdieu's use of habitus in his theoretical scheme. McLeod argues that "the circular relationship Bourdieu posits between objective opportunities and subjective hopes is incompatible with the findings [presented here]" (pg. 355). The Brothers probably had even lower objective opportunities than the Hallway Hangers as a result of race, but they still held on to greater aspirations. McLeod sees Bourdieu's failure as the inability to understand the number of factors that mediate the effect of social class on individuals. What McLeod's work shows is that the ways in which individuals respond to the structures of domination that they face are innumerable. What we learn from the Hallway Hangers and the Brothers is that their lives cannot be reduced to structural influences and causes. McLeod quotes Willis, we must "'give the social agents involved some meaningful scope for viewing, inhabiting, and constructing their own world in a way which is recognizably human and not theoretically reductive'" (pg. 356).

The "New Structuralism"

Michael J. Piore - The Dual Labor Market: Theory and Implications"

Piore's argument is that "the role of employment and of the disposition of manpower in perpetuating poverty can best be understood in terms of a dual labor market" (pg. 359). Understandings of the labor market have tended to consider it as a uniform system. Piore (and others to come) argue that this is not true. In fact, there are two sectors of the labor market, the primary and the secondary, that are organized in very different ways. The primary sector offers jobs that have high wages, good working conditions, employment stability, chances for advancement, and equity and due process in the administration of rules. In contrast, the secondary sector offers jobs that have low wages, poor working conditions, a lot of variability in employment, harsh and often arbitrary discipline, and little opportunity to advance. The secondary sector is where the poor are confined. Piore argues that to eliminate poverty, they must gain access to primary employment.

Piore lists several factors that confine the poor to secondary positions and maintain the duality of the labor market.

1. The most important characteristic separating primary and secondary jobs are the behavioral requirements. He argues that the main thing that stands in the way of secondary workers being successful in primary positions is their inability to show up for work regularly and on time. These qualities are much less significant in secondary employment and employers are much more tolerant.

2. Secondary workers who have the behavioral traits needed for primary work are trapped in the secondary market by superficial characteristics that liken then to other secondary workers (like race, accent, education, demeanor, etc). This happens because employment decisions are usually based on a few readily observable traits (like race, accent, education, demeanor, etc.). So the qualified persons are victims of statistical or pure discrimination. Discrimination increases the secondary labor force thus lowering the wages secondary employers must pay. So secondary employers have an interest in perpetuating it, as do primary employees because it limits the labor supply in that sector and raises wages.

3. The distinction between primary and secondary jobs does not appear to be determined by technology. A lot of work can be organized either for stable or unstable workers.

4. The process of working in the secondary labor force and living among secondary workers reinforces the behavioral traits associated with the secondary sector.

5. Among the poor, income sources besides employment, such as public aid and illegal behavior, are more compatible with secondary work than with primary work.

The central implications of the dual market interpretation of poverty: the poor do participate in the economy, the manner of their participation is the problem. "Their current mode of participation is ultimately a response to a series of pressures-economic, social, and technical-playing upon individuals and labor market institutions" (pg. 360). Regarding manpower policy, this means that those focused on training, counseling and placement for the poor are not very effective because they are simply attempts to move people from the secondary to the primary sector. They do not attack the system itself. "More weight should be placed upon policies which affect the environment in which employment decisions are made and the pressures which the environment generates" (pg. 361). Such policies include antidiscrimination policy, occupational licensing reform, and the structure of public assistance.

Finally, we must recognize that there are groups who are actively interested in perpetuating poverty (for example, remember the groups who had an interest in discrimination?). Policies that attempt to move people out of the secondary sector are in danger of subversion by these groups. The trick to successful policies is to find a way to avoid both rejection and subversion.

Sorensen and Kalleberg - "An Outline of a Theory of the Matching of Persons to Jobs"

S&K begin with a brief critique of neoclassical or orthodox economic theory (Neo) regarding labor market processes. This theory wage determination uses marginal productivity theory to explain the demand side and human capital theory to explain the supply side. They argue that sociologists do not need a separate theory if the Neo approach accounts for the empirical findings well. However, there are several problems with it. Many of its predictions about the earnings attainment process, labor supply and more have not been supported. Most problematic is the assumption that the labor market is homogenous in nature. Critics argue that jobs and jobs structures vary (i.e. the dual labor market as Piore discussed). S&K present a conceptual framework for analyzing labor markets. "Labor markets are arenas for the matching of persons to jobs. The conditions that determine the earnings outcome of this matching process are of primary interest here, particularly the identification of what determines the influence of job and personal characteristics on earnings" (pg. 363). Their intent is not to prove the Neo theory wrong, but to identify the conditions in which it does and does not work. They argue that the conditions necessary for the matching process described by Neo are absent in some segments of the labor market. They argue that alternative matching processes are found in such segments which have very different implications for the earnings determination process and other labor market processes.

S&K use Weber's concept of open and closed social relationships to identify different job structures with different matching processes. The degree of closure is determined by the bargaining power of employees and employers. They focus on the consequences of variation in control over the job by the employee vs. the employer. Two dimensions of control are considered: control over the activities of the job (i.e. employee autonomy or lack thereof) and control over access to the job. The two dimensions may vary independently and control over access is considered to be especially crucial because it affects the competition between workers. The degree of control over access is a continuum. At the "closed" extreme, the employee "owns" the job and cannot be involuntarily replaced. At the "open" end, the employee is at the mercy of the employer and can be replaced at any time. S&K's main argument is "that the mechanisms the employer can use to relate wages to performance depend on the employment relationship, particularly the employees control over access to the job, and that these different mechanisms identify important differences in labor market structures relevant also for labor market processes other than earnings" (pg. 364).

Briefly, the important proposition of Neo is that earnings differentials reflect differences in the productive capacities of people as a result of their training, abilities, and training opportunities. The labor market is assumed to be competitive and perfectly functioning, in addition to being homogenous. S&K argue that this 'wage competition model' only occurs when employment relationships are open. Closed employment relationships protect incumbent workers from competition. They are not arguing that closed relationships result in inefficient production. They are saying that the promotion systems involved represent very different mechanisms for attaching wages to performance than the wage competition model. In closed systems, "the wages for individual employees will reflect the jobs they hold and therefore, not only their performance, but also the rate at which vacancies appear, the organization of jobs, and the seniority of employees" (pg. 366). This is a very different labor market structure than the one depicted by Neo.

S&K refer to the matching process in these closed relationships as vacancy competition. Since employees have control over access to the job, a vacancy must exist for someone to get access. This is not the only alternative to wage competition (which occurs in open relationships), but it is the one they are most concerned with here. In vacancy competition, employers can no longer link marginal productivity to wage because of the indeterminate length of employment. This has important consequences for hiring decisions, wage determination, and the organization of jobs in hierarchies.

In hiring, the employer must consider the potential performance of the candidate and the relationship between personal characteristics and productive capacity because once hired, the employee cannot easily be fired. So, there will be a ranking of candidates. Regarding wages, wage rates in vacancy competition are characteristics of jobs, not a measure of the productivity of the person. Job ladders create incentives, but must be carefully organized so that jobs at the same level provide the same earnings and the differences between levels must be significant enough to encourage competition between employees for promotion. Promotion schedules also encourage low-performance employees to leave voluntarily so they can be replaced, hopefully, with more productive individuals. Promotional opportunities are created when someone leaves or a new job created and the individual career appears as a succession of movements into and out of positions. S&K say that these promotion schedules "will under certain conditions result in career lives that are similar to those predicted by human capital theory, even though the mechanisms are quite different" (pg. 367).

So, in vacancy competition, variations in earnings reflect variations in job characteristics and the organization of jobs in internal labor markets rather than reflect variations in personal characteristics that determine the productivity of the individual, as in wage competition. These two opposite models of the matching process have different implications for a person's ability to increase her earnings. In wage competition, inequality lies outside the labor market, in education and training. In vacancy competition, the major constraint on income attainment is access to the jobs. Inequality is a result of restricted access to jobs. S&K argue that vacancy competition is predominant in the U.S. economy. Hence, the very limited success of worker training programs.

Mark Granovetter - "Toward a Sociological Theory of Income Differences"

In this article, G is not proposing a theory. Rather, he is mulling over the sociological and economic approaches to income differences and trying to pull out what seems useful. He hopes this will lead to a reconstruction and integration of the two.

He argues that, in modern economic systems, earnings are determined by: 1) personal characteristics (prestige goes here); 2) job characteristics ( the idea that jobs exist independent of their incumbents (as Weber argues); 3) processes that match 1 and 2 (how people with certain personal characteristics are matched to certain jobs. He categorizes existing theories of income differences by how much they emphasize each of these. An adequate and useful theory needs to consider and integrate all three, but existing theories can be characterized as focusing almost completely on one.

He reviews these theories based on this categorization:

1. Theories that focus on personal characteristics - status attainment research and human capital theory. Previous discussions of these should have been sufficient. It is obvious that these two approaches focus on the individual's characteristics. If you want to review, the discussion is located on pp. 372-374. G. notes that the problem with these approaches is the lace of attention given to employers, matching processes, and the mechanisms by which investments generate a stream of income.

2. Theories that focus on job characteristics - Institutional economics: wage structures, segmented labor markets, and labor queues. The concept of 'wage structure' (introduced by Dunlop) directed attention to how a particular job came to have the wage associated with it and how these jobs fit into a hierarchy. So that, for ex., a wage increase/decrease in one position in the wage structure will set off a chain reaction to other positions surrounding it.

Next we come back to the segmented (an extension of the dual) labor market which Piore discussed. G. notes that some attention is supposed to be devoted to how people are placed in the different sectors, especially how people get stuck in the secondary sector, but in practice little attention to matching processes is paid. "Instead they argue that confinement occurs because workers lack stable work habits. G is disappointed in this reductionism, I think, as was I. He also finds the theory static and atomistic.

The idea of a labor queue was introduced by Thurow. Job candidates are supposed to be ranked by employers according to their trainability. If supply of qualified candidates is greater than demand, then some workers will end up in worse jobs than the 'deserve'. G. finds this approach empirically difficult to distinguish from human capital and status attainment because 'trainability' depends primarily on background characteristics and educational achievements. Also, it still does not take the matching process seriously enough for G.

3. Theories of matching job and personal characteristics - G's problem with these theories is that they only address the demand side or the supply side, but not both. He discusses approaches to employers' 'signaling' or 'screening' on the demand side and employee job search on the supply side. "Such theories ought properly to apply to both sides of the market. Employers as well as employees search, and employees as well as employers try to read signals from the other side of the market. The one-sidedness of both sets of ideas precludes sophisticated approaches to the question of how employers and employees are matched. Furthermore, an adequate theory would have to incorporate both searching and signaling, whereas present theories consider these as sequential activities" (pg. 379).

Finally, G. notes some limitations to his argument here. First, the omission of a discussion of discrimination which is an important element of the matching process. He also points out that his three part classification is a useful initial probe, but has limitations. It is somewhat atomistic and ignores the structure of the organizations within which the matching occurs which most certainly impacts the process. Finally, he notes the omission of a discussion of the impact of macroeconomic forces on income differences.

Concluding Commentary to Part IV

James N. Baron - "Reflections on Recent Generations of Mobility Research"

Baron says that the articles in this section demonstrate why mobility and attainment research has been so influential. He quotes Featherman as saying that the subfield is virtually unmatched in its cumulative development of theory, models, methods, and data. In this commentary he addresses the contrast between 'structuralist' and 'individualistic' perspectives. The former, he argues, is interested in differentiating between contexts, the latter in differentiating between individuals. B thinks that "the whole contrast between so-called structuralist and individualistic perspectives may have been either misunderstood or overdrawn" (pg. 385). In this essay, he points to some ways that future research could profit by considering issues emphasized by the structuralist approach.

Orthodox mobility and attainment studies have been criticized as being inattentive to the structural issues. Baron argues that recent work has begun to address macro forces influencing stratification and is not as individual-oriented as earlier work. He also does not much like the classification of classical mobility and status attainment approaches as individualistic. He argues that the Parsonsian framework underlies much of this work means that there is an assumption of functionality in the larger societal level. Baron notes that critics of the functionalist nature of the research were 'right for the wrong reasons'. "That research is functionalist not because it employs occupational status scores or analyzes individual-level data but because of its assumption about the macro forces shaping the allocation of personnel in postindustrial societies" (pg. 386).

The new structuralists, B says, have overpromised and underdelivered. Three observations: 1)despite increasing recognition of the important role of organizations, little explicit study has been undertaken; 2) stratification and mobility researchers tend to infer organizational and institutional mechanisms from wage and mobility patterns rather than observing them directly; 3) the new structuralist stories are surprisingly economistic, there is seldom anything distinctly sociological in them. Baron argues that the structuralists can be criticized on the flip side of the criticisms they make: it is difficult to discern individuals and social interaction in their accounts.

So where do we go from here? Baron argues that "the conceptual framework underlying status-attainment research is 'structural' in a sociological sense, whether or not the analyses have been 'structural' in the statistical sense. In other words, the trends and relationships among variables observed in attainment studies are seen to reflect fundamental transformations in the linkages among the institutions of family, education, work, and the polity in advanced economies rather than, say, innate differences" (pg. 388). He goes on to suggest scattered avenues for future investigation. He suggests that the focus on global dimension of stratification, such as income and occupational status, has been at the cost of ignoring local dimensions of status and social/psych factors influencing aspirations, decisions, and satisfactions. The preoccupation with occupational classification schemes has drawn attention away from the ways roles are defined in particular settings. How work activities get classified into job categories depends on who is doing the work and who is doing the classification (for instance, gender differences in titles for the same position). B also discusses ways in which the role of occupations as the building blocks of stratification may vary across time and space. He argues that social and economic developments may be weakening the link between occupations and lifestyles.

What all this boils down to is this: B thinks that mobility and attainment researchers have been very successful in measuring 'surface phenomena', but have been much less successful in refining their theories about the 'deep structures' that generate those phenomena. It is on these 'deep structures' that we need to focus serious attention and work on developing an understanding of.