Social Structure and Mobility in Economic Development

Seymour Lipset, Neil J. Smelser, Seymour Martin Lipset

• Publisher: Routledge
• Year: 2018
• eBook ISBN: 9781351306225
• Topic: Économie
• Subtopic: Économie du développement

The foundation of this volume is the notion that the several processes of change constituting economic and social development are systematically interrelated. The essence of development is the appearance of rapid rates of increases in many different indices--output per capita, political participation, literacy and the like. These quantitative changes are, however, commonly accompanied by vast changes in the social structure--markets emerge, political bureaucracies arise, and new educational systems appear. Written by the leading authorities on the subject, this group of papers tackles the causes and consequences of social mobility. Each author brings his particular skills to bear on various aspects of the problem in studies of persons moving from rural to urban settings, from one kind of industry to another and from one prestige level to another. Several of the papers review the theoretical and methodological issues involved in comparative research on social mobility while others compare and contrast traditional and modern stratification systems. Various papers explore the economic, religious and psychological basis of social mobility, concluding with enquiry into the consequences of rapid mobility, especially in terms of the political stability of developing nations. Because social mobility is a central consideration in any study of economic and social change, every student of change will use this pioneering reference source as a text for all future research. Contributors include Otis Dudley Duncan, Harold L. Wilensky, Michael G. Smith, Bert F. Hoselitz, Wilbert E. Moore, Natalie Rogoff Rams°y, Gideon Sjoberg, Reinhard Bendix, Harry Crockett, David Matza, Lester Seligman, and Gino Germani. Neil J. Smelser is emeritus professor, Department of Sociology, University of California, Berkeley. Seymour Martin Lipset was professor of sociology and director of the Institute of International Studies at the University of California, Berkeley.

2

Methodological Issues in the Analysis of Social Mobility*

OTIS DUDLEY DUNCAN, University of Michigan

THE JUXTAPOSITION of themes—social mobility and economic development—in the title of the conference may invite acceptance of an unwarranted assumption. In point of fact there is and can be no fixed and determinate general relationship between measures of economic growth and indexes of social mobility, either over time in one country or between countries at a point in time. A whole set of auxiliary postulates, each empirically contingent, must be adopted before a relation between mobility and growth can be deduced.

The work of Kuznets, Clark and others has indeed supported the proposition that economic growth, in the sense of sustained increase in output per capita, is accompanied by a redistribution of the working force by functional categories—industries or occupations. Yet, such a proposition, even if accepted as an axiom, implies very little about the kinds, amounts, and patterns of social mobility that will be observed during a sequence of economic growth.

To see why this must be so, consider observations on an economy taken at two times, t₀ and t₁. Let the functional distribution of the working force be described as the frequency of employment in each of k classes (including, possibly, null frequencies for some classes). If we designate the distribution at t₀ as a vector u₀ and that at t₁ as a vector u₁, the issue of social mobility arises in trying to account for the transformation of u₀ into u₁. Evidently, such a transformation includes the following component processes of “social metabolism:” exits from u₀ produced by death, migration out of the country, and retirement from economic activity; entries into u₁ produced by initiation of gainful employment or migration into the country; and, for personnel included in both distributions, change of functional class, i.e., mobility. Note that mobility is but one of the components of the transformation. Observe, moreover, that mobility from, say, class i at t₀ to class j at t₁ may be balanced by mobility from class j at t₀ to class i at t₁ Thus, even if mobility were the only component, our axiom only tells us that a net redistribution of the working force must occur, while a given net redistribution can be accomplished by an indefinitely large number of patterns of gross mobility. As soon as the other components are allowed, it is evident that any redistribution whatsoever can be accomplished by them alone, in the absence of any mobility whatsoever; or, on the other hand, the redistribution produced by mobility as such can be compensated or amplified, in any manner or degree you care to specify, by the other components.

In the foregoing account the only type of mobility considered is the change of occupational position on the part of a constant set of individuals, i.e., intracohort or so-called intragenerational mobility. If we introduce the possibility that an individual’s position at either t₀ or t₁ depends in some fashion on his “origin” (say, his father’s occupation) the analysis is complicated and the possibility of a determinate relation between economic growth and mobility pattern is, if anything, reduced. Foremost among the complications is the fact that a distribution of the fathers of those in the working force at a given time is not, and in the nature of the case cannot be, a distribution of the working force at some definite prior instant of time. This follows from the facts (a) that fathers vary in the number of sons they may have, from zero to some not well-defined upper limit; and (b) that age at paternity is a variable, i.e., a father may be no more than 20 years (or even less) older than his son, or as much as 50 years (or even more) older. Ordinarily, in the conventional study of intergenerational mobility, the origin distribution represents fathers in proportion to the number of their sons of a given age, and aggregates into a single distribution the positions they occupied over a considerable period of time.

Another way to express the predicament is this. Presumably, there is an actual, though unknown, transformation T which takes into account all the sources of change in an occupation structure, so that u₀T = u₁. We could then write u₀ = u₁T⁻¹. But suppose a transformation S exists such that u₀ = u₁S. Can we then infer that S = T⁻¹ and accordingly T = S⁻¹, yielding a solution for T? Unfortunately, no. That is, to discover a transformation that could have sent u₀ into u₁ is not necessarily to learn what transformation actually did that job.

The upshot of these considerations is that the connection, if any, between economic growth and social mobility falls into the domain of contingent rather than necessary relationships. To be sure, some of the contingency may be removed by introducing assumptions that restrict, or even eliminate, certain of the components of occupational redistribution. Thus, for purposes of constructing a model of social mobility the analyst may choose to consider a population closed to migration. But this is to abstract from the real process of economic growth as well as the real process of occupational mobility. While partial deductive models, therefore, have their analytic uses, the generalizations reached concerning interrelations of mobility and economic development must be in large part empirical generalizations, supported or rationalized as well as they may be by some not very specific deductions from apparently applicable postulates.

This situation has already been encountered in the paper by Smelser and Lipset, wherein they refer to the logic that economic growth should result in a pattern of high upward and low downward mobility, whilst noting that the “logic” is not always borne out by the data. Apparently, the “logic” rests on one or more assumptions that are contrary to fact. They themselves, at this juncture, call for a resort to inductive procedure: “What is necessary is a systematic effort to relate the data of changes, classified in different ways, to the estimates of growth.”

These remarks suggest, among other things, that a sharp separation of issues in “theory” from problems of “method” is no more justified in the study of social mobility than in any other domain of inquiry. The methodologist is concerned with the grounds for accepting bodies of data as evidence for or against propositions. If he can show how verbal formulations place logically irreconcilable demands on bodies of data—those in existence, or those that might conceivably come into existence—then revision of the formulations is in order. Rather than “answering” the questions posed by theory, fundamental study of analytical methods is likely to make two questions grow where only one flourished before.

This paper makes two kinds of contribution. In the next section, the issue of how occupational mobility may be related to changes in occupational structure is considered in greater detail. Some positive results are obtained, but the topic is left at the point where the “theorist” must vouchsafe a revised formulation as a basis for further methodological and empirical explorations. The remaining sections of the paper deal with a selection of technical and methodological issues raised in the literature on intergenerational occupational mobility. The issues are ones I consider important and on which I profess to have something to say. They hardly exhaust the agenda of pressing problems. The vexing topic of what effect errors in mobility data have on conclusions, treated cursorily in the first draft of this paper, is omitted for lack of space. Other matters meriting discussion but ignored here are covered in a recent methodological paper by Yasuda,¹ with most of whose points I am in general agreement, despite some reservations about specific procedures.

Occupational Mobility and the Transformation of Occupational Structures

If the reader, like the writer, finds it helpful to have the problem posed concretely rather than in purely abstract terms, he may refer to Table 1, where occupational distributions for two years are exhibited in columns (a) and (b). The years, 1960 and 1930, were selected so as to be about a “generation” apart. (The median age of all men who have sons born in a given year usually is around 30 or 31.) The net changes in the male occupational distribution in the U.S. during this period are matters of common knowledge—the rapid increases in proportions of white-collar, particularly professional and technical jobs; the likewise appreciable increases of occupations involving skilled or semi-skilled manual work; the marked decline in proportion of laborers; and the even more pronounced decrease in percentage of farm workers, both farm operators and farm laborers. Over the period in question, moreover, the absolute size of the male labor force increased by about one-fourth.

The character of the transformation that produced these changes is not so well understood. Indeed, the information needed for a complete analysis of it is not now and may never become available. We can, nevertheless, profitably study the problem of whether certain components of the transformation could, in theory, be related to the idea of occupational mobility.

A convenient point of departure is the well-known discussion by Kahl,² which is noteworthy as an explicit formulation of the problem, superseding a mass of prior literature. Although the conclusion of the present discussion is that Kahl’s effort was a failure—and for fundamental reasons, not merely because of flaws in the data available to him—i...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Contents
• AldineTransaction Introduction
• Preface
• I Social Structure, Mobility and Development
• II Methodological Issues in the Analysis of Social Mobility
• III Measures and Effects of Mobility
• IV Pre-Industrial Stratification Systems
• V Interaction between Industrial and Pre-Industrial Stratification Systems
• VI Changes in Occupational Structures
• VII Changes in Rates and Forms of Mobility
• VIII Rural-Urban Balance and Models of Economic Development
• IX A Case Study in Cultural and Educational Mobility: Japan and the Protestant Ethic
• X Psychological Origins of Mobility
• XI The Disreputable Poor
• XII Political Mobility and Development
• XIII Social and Political Consequences of Mobility
• Index