The Deferred Gratification Pattern. A Preliminary Study (Schneider L. - Lysgaard S., 1953)
The Deferred Gratification Pattern. A Preliminary Study (Schneider L. - Lysgaard S., 1953)
The Deferred Gratification Pattern. A Preliminary Study (Schneider L. - Lysgaard S., 1953)
Review, Vol. 18, No. 2 (Apr., 1953), pp. 142-149 Published by: American Sociological Association Stable URL: http://www.jstor.org/stable/2087723 . Accessed: 07/10/2013 22:09
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to American Sociological Review.
http://www.jstor.org
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
tinent current literature emphasizes, by way of example, middle-class renunciation of impulses toward violence. The concepts of deferment and renunciation tend, however, to overlap. One may renounceonly temporarily, and the question as to whether we are dealing with renunciation or with deferment becomes verbal. A more important point is the normative character of the deferred gratification pattern. Middle class persons feel that they should save, postpone, and renounce a variety of gratifications. There are very probably also normative elements in the "lower class" pattern of non-deferment. Thus, Whyte notes that one of the important divergences between the social mobility pattern and the corner-boy activity pattern in Cornerville appears in matters involving expenditure of money. The college boys save money for educational purposes or to launch business or professional careers. But the corner boys must share their money with others and avoid middle class thrift. Should a corner boy have money and his friend not have it, he is expected to spend for both. The corner boy may be thrifty, but, if so, he cannot hope to hold a high position in the corner gang.3 Through the work of Kinsey and his associates much clarification has been given to the deferred gratificationpattern as it manifests itself in the sexual sphere. Ginzberg notes that "the poor" discount the future more heavily than "the better situated classes," and thereby, in our terms, as far as sexual intercourse is concerned, they do not defer gratification, i.e., relatively to "higher"classes.4 Kinsey and his co-workers
cal literature. See, e.g., A. Davis and R. J. Havighurst, "Social Class and Color Differences in Child
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
THE DEFERRED GRATIFICATION PATTERN supply pertinent evidence.5 Relatively, "lower" class persons indulge considerably in pre-maritalintercourse; "upper"class persons show relative deferment of gratification in this sector of behavior. One illustrative commentmay be remarked.This is to the effect that in two or three lower level communities it was not possible to discover a single male who had not had sexual relations with girls by the time he had reached the age of sixteen or seventeen. The rare boy who had not had such relations by that age was either physically handicapped, mentally defective, homosexual, or "earmarked for moving out of his community and going to college."6 Work designed to describe the pattern of deferred gratification in something like its entirety is especially marked in some of the research on Negro classes. Drake and Cayton, in probing upper-class definitions of lower class life, emphasize the definition of a pattern of that life. They summarize some of the testimony of upper class interviewees, indicating that a few of those interviewed defined the lower class solely in terms of economic criteria, high income meaning high status, and low income low status; but that a larger number thought of a pattern, a constellation of traits, when they sought a definition of the lower class. The trait that most consistently emerged as clue to lower class status was "rowdy or indecorous behavior in public." Lower class people do not restrain their emotions and are ignorant of "how to act," of correct dress, of wise expenditure of money.7 One gathers from the context that Drake and Cayton would largely accept the descriptive, if not the evaluative,
143
5 Cf. the data of A. C. Kinsey, W. B. Pomeroy and C. E. Martin, Sexual Behavior in the Human Male, Philadelphia and London: W. B. SaundersCo., 1948, Ch. 10. 6 Ibid., p. 381. The words quoted point to what may become one of the most significant foci of study of the deferredgratificationpattern-the extent to which there is early adoption by those who of the class patterns of the become upward-mobile "higher"classes,a problemgiven some emphasisin the work of the Warnerschool. 7 St. Clair Drake and H. R. Cayton, Black Metropolis,New York: Harcourt, Brace and Co., 1945, pp. 562-3. See also, ibid., pp. 559-562. Pertinent statementsor incidentalcommentsbearingdi8 A. Davis and J. Dollard, Childrenof Bondage, rectly or indirectly upon the deferred gratification pattern or its absence,will be found, ibid. pp. 523, Washington,D. C.: American Council on Education, 1940. 586, 590, 592, 608, 661-2, 690, 692, 705, 714.
aspects of this upper class view of the lower class pattern of life. The relevance of the description to the deferred gratification pattern is unmistakable. An earlier study by Davis and Dollard,8likewise of classes within a Negro community, is also highly pertinent and constituted perhaps the most immediate inspiration to the undertakingof the research reported in this paper. The present writers proceeded on the assumption that the Davis-Dollard study might be illuminating for (at least American) class structure in general, despite its formal concern with classes in the Negro community alone. The study indicated that "impulse-following" (with minimal deferment or renunciation of "impulses" or gratifications) was characteristic of lower class Negroes and "impulserenunciation" (or deferment of impulsegratification) characteristic of middle class Negroes in "Old City" in the South. The study is more insightful than rigorous, and there are undoubtedly difficulties with its categorizations of class itself. Nevertheless, we may construct from it a useful listing under the general caption of the class patterning of deferred gratification.This listing served as a starting-point in our own research. The lower-class-characteristic "impulse-following" (absence of deferred gratification pattern) involves: relative readiness to engage in physical violence, free sexual expression (as through intercourse), minimum pursuit of education, low aspiration level, failure of parents to identify the class of their children's playmates, free spending, little emphasis on being "well-manneredand obedient," and short-time dependence on parents. On the other hand, the middle-classcharacteristic "impulse-renunciation"(presence of deferred gratification pattern) involves the reverse of these traits: relative reluctance to engage in physical violence, and the like. Admittedly, this is a rough listing and leaves us still far from a thoroughly comprehensive specification of the deferred gratification pattern. Some of the items may appear equivocal. For example, identification of class of playmates by parents perhaps does not carry on the face of it the reason
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
144
AMERICAN SOCIOLOGICALREVIEW
In an investigation dealing with differential responses among social classes, it is, of course, of particular importance that our sample is more homogeneously middle class than the total population of high school age. The homogeneity of the sample produces an underestimate of the class differentials in the population at large. The sample does not extend adequately into the lower rungs of the stratification ladder. This probable under-representation of the lower class should serve to increase confidence in the trends actually observed. Subjects were asked to check one of twelve occupational classes, designating the one most similar to the occupation of their fathers. These twelve occupational classifications were grouped into four occupational classes according to degree of supervisory power over "lower" occupations and independence of supervisory control from "higher" occupations. The four occupational classes follow: Class 1. Independent Occupations (including executives, directors, owners of business or farm, doctors, lawyers, bankers, ministers, professors, consulting engineers); Classi2. Dependent Occupations Involving Skill and Supervision or Manipulation of Others (including supervisors, foremen, technical engineers in industrial employment, sales workers, agents, clerks or secretaries in small businesses, teachers, nurses, preachers, reporters, public officials, entertainers); Class 3. Dependent OccupaTHE RESEARCH Involving Skill but Little Supervision The data utilized for the present research tions or of Others (including skilled Manipulation are based on a representative sample of and semi-skilled workers and workers in 2500 high school students. This is drawn industrial employment); Class 4. Dependent from a larger nation-wide sample of some OccupationsInvolving Little Skill and Little 15,000 high school students.10Although the Supervision or Manipulation of Others (insample, as stated, is "representative," it is cluding assembly line workers, laborers, not representative of all the nation's teenjanitors, farm workers, road workers, miners, agers, only of the high school population. drivers). These occupational classes are admittedly 9Ibid., pp. 119 and 133. not internally homogeneous in respect to 10 The data were obtained through the Purdue such class criteria as "income," "prestige" Opinion Panel, Poll Number 32, administered by the Division of Educational Reference at Purdue or "social equality." It was intended that University, Indiana. The Purdue nation-wide samthe classification be independent of such ple of 15,000 is fairly representative of the total which at their face value might be criteria, high school population in the country in such faccontingent upon a certain way of life. The tors as sex, rural-urban residence, religion, political variable, the deferred gratification pattern affiliation, level of education of mothers, and house and home characteristics. The smaller sample of (DGP), which is to be related to class 2500 which we utilize has been stratified according is itself at least an important affiliation to geographical region and grade in school, but of a "way of life," and we wished element return of 15,000 strictly randomized from the total to avoid as far as possible the danger of with respect to the other characteristics listed above. for its inclusion. But it is conceived that such identification implies a concern for the maintenance of certain standards and a fear lest they be jeopardized by the contamination of "unfortunate" associations. Some items mentioned by Davis and Dollard are omitted from the list. Thus, reference is made to the "reliable middle class conscience" by contrast with the "less reliable" lower class conscience.9 This is at a somewhat different level of abstraction from the other traits and activities noted and not so easy to handle in the type of research to be reported, and it has therefore been omitted. Finally, following Davis and Dollard, the listing tends to focus on children, but the implications plainly extend beyond them alone. Although nothing like exhaustive documentation of the consideration of the deferred gratification pattern by sociologists has been attempted, and although we have not sought to offer anything like a finished specification of the pattern itself, the above is perhaps sufficient to suggest that the pattern deserves systematic investigation and that it is time to co-ordinate the bits of researchendeavor that have been directed toward its study. The following preliminary and partially reported research is designed to do no more than make an appropriate step forward.
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
THE DEFERRED GRATIFICATION PATTERN obtaining an association between "class" and "response" that might turn out to be in considerable part tautological." The writers have emphasized that the occupational "classes" are not "social classes" in the sense of comprising people with clearly "common" interests, or with "class consciousness." In order to get a more "subjective" classification of our subjects we also asked them to choose one of four class designations as the one most fitting for themselves and their families: the "upper class," the "middle class," the "working class," and the "lower class."'12
145
The poll included twenty-eight DGPrelated questions. It is impossible to present all of the results here or to give them full statistical exploration. We can only affirm that our results generally give good support to the hypothesis that a classrelated DGP pattern exists, especially when the self-identification index of class is used. Beyond this mere affirmation, we can present some selected results that are of considerable interest. Fairly typical results emerge if we follow the listing of patterns described by Davis and Dollard, and the discussion following is confined to these,. Physical Violence. The poll results show OF FOUR TABLE1. CLASSSELF-IDENTIFICATION that the students who identify themselves CATEGORIES* OCCUPATIONAL with the "working class" report in a slightly higher proportion than those who identify Self-Identification Occupational themselves with the "middle class" that they Total Classes "Middle" "Working" have had one or more fights recently (dif100 20 80 1. Independent ferences significant at the 5 per cent level), 2. Dependent, that they have seen adult fights recently 25 100 75 manipulative (differences not significant at the 5 per 100 63 37 3. Dependent, skilled cent level), and that they prefer to "settle 100 50 4. Dependent, unskilled 50 matters right away" rather than "let their * Figures in percentages. Differences are signifitemper quiet down" first (differences sigcant beyond the 1 per cent level. nificant at the 1 per cent level). Slight differences are also found among the occuIt proved feasible, however, to use only classes in respect to the same quespational "middle" and "working" class self-identifithe lower occupational classes giving tions, cations, assimilating the very few other the above responses more frequently than responses to these. the higher ones. The magnitude of the differences is in no case impressive, however, 11 This points to a danger that, in general, conand we consider that these results give only stantly besets current class research. For example, slight support to the Davis-Dollard obserHollingshead's proposition that there is a functional relation between the class position of the family of vations, as far as the occupational differenan adolescent and his "social behavior" in his comtiations go. munity (Elmtown's Youth, p. 441) would be more Free Sexuality. The polling technique is impressive if the rating procedure Hollingshead originally relies upon did not allow for the possinot the best one for studying the sexual bility that the adolescent's social behavior itself behavior of high school children. The poll enters as an element into the determination of the questions had to be formulated rather class position of his family. because of prevalent attitudes revaguely 12 The procedure is borrowed from Richard Centers,The Psychology of Social Classes,Prince- garding sexual matters in our high schools, ton: Princeton University Press, 1949, Ch. VI. The and they failed to yield any clues to class middle class bias of our sample is indicated by the in sexual behavior. The writers differences high frequency of "middle class" self-identification believe this failure to be accidental; it is (about two-thirds of the subjects so identifying certainly not definitive. themselves), compared to Centers' figure for his national, male, white sample (less than one-half of Marked Pursuit of Education. Data rethe subjects identifying themselves as "middle to education are presented in Tables lating class"). The relationship between the two classifica2 and 3. tions may be of some interest and is shown. (Since very few of our subjects marked "lower class" and The limitations of the data of Tables 2 "upper class" as their choice, the few "lower class" and 3 are evident and need no special comresponses have been added to the "working class" ment. The results are nevertheless in concategory, and the "upper class" responses to the formity with the DGP hypothesis, although "middle class" category.)
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
146
AMERICAN SOCIOLOGICALREVIEW
TABLE 2. PLANS AFTER GRADUATION FROM HIGH SCHOOL OF STUDENTS FROM DIFFERENT SOCIAL CLASSES*
Plans To go to college or take special training other than college To go to work, enlist in military service, or "other plans"
Total
56 42
44 58
100 100
62 61 48 32
38 39 52 68
* Figures in percentages. The chi square test shows that differences are significant beyond the 1 per cent level for both class indices.
obviously they do not give conclusive proof of it.13 High Aspiration Level. Students were asked on a list of twelve occupations (ranging from high to low in terms of income
13 In the original,more completestudy, the data were analyzedfor partial association,i.e., the association between class of self-identificationand response,within each of the occupationalclasses,and the associationbetween occupationalclass and response, within each of the classes of self-identification. This analysis showed that when the variation class is accountedfor there is still a of occupational significantassociationbetween class of self-identification and response.Thus, we may say that "ecoas far as that may be reflected nomic opportunity," by occupationalclass, is not solely responsiblefor among the classes. the differences
and prestige) to check those which they considered "not good enough" for their own life work. The higher the number thus checked the higher the occupational aspirations of the students. Table 4 summarizes relevant data. Identification of Class of Playmates by Parents. On the item, "My parents have definite ideas about what my friends should be like," there are small differences (not significant at the 5 per cent level) among the classes, the parents in the higher classes being reported to be more concerned about friends of children. The students themselves, however, seem to differ in respect to concern with the family backgrounds of their friends. Thus, 66 per cent of those
OF DIFERENT SOCIAL AMONG FROmHIGH SCHOOL STUDENTS OF GRADUATION TABLE 3. EXPECTATION CLASSES* Expectations of Graduation Definitely expect to graduate Do not expect to graduate or will only "probably" graduate
Total
80 70
20 30
100 100
82 82 76 68
18 18 24 32
* Figures in percentages. The chi square test shows that differences are significant beyond the 1 per cent level for both class indices.
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
147
5.4 4.8
Responses to statement, "In my family we always seem to be broke just before payday, no matter how much money is coming in." Class Index Self -Identification "Middle" "Working" Occupational Class 1 2 3 4 16 32 39 45 84 68 61 55 100 100 100 100 28 43 72 57 100 100 Agree Disagree Total
* The difference between the medians is significant at the 1 per cent level in the case of the classes of self-identification. The median for the lowest occupational class is smaller than any of the medians of the other classes by a statistically significant amount. The differences of the medians for the upper occupational classes are not statistically significant.
who identify themselves with the "middle" * Figures in percentages. The chi square test class and 56 per cent of those who identify shows differences are significant beyond the 1 per themselves with the "working" class (dif- cent level for both class indices. ferences significant at the 1 per cent level) say that they "enjoy being together with Being Well-Manneredand Obedient. This friends who come from families at least as trait, or complex of traits, was examined nice and successful as my own, rather than only in the specific matter of table manners. being together with just anybody."'4 Table 7 shows the pertinent data for the classes of self-identification but omits the 14 On this same item, there is an opposite, aldata for occupationalclasses, for which there tbough not statistically significant, tendency among was no clear trend. the occupational groups. Prolonged Depeitdence on Parents. This AMONG HIGH PREFERENCES TABLE5. SPENDING was examined through the single indicator OF DIFERENT SOCIAL STUDENTS SCHOOL
CLASSES* Responses to the question "If you won a big prize, say two thousand dollars, what would you do?" Spend most of it right away Save most of it
TABLE 7. CONCERN FOR TABLE MANNERS AMONG PARENTS OF HIGH SCHOOL STUDENTS FROM DIFFERENT SOCIAL CLASSES*
Total
Responses to question, "How would your parents feel if you rested your elbows on the table while eating and talked with your mouth full?" They would object They would object a little or wouldn't mind
27 32
73 68
Total
100 77 23 1 100 72 28 2 100 69 31 3 100 66 34 4 * Figures in percentages. The chi square test shows differences are significant beyond the 1 per cent level for both class indices.
74 67
26 33
100 100
* Figures in percentages. The chi square test shows differences are significant beyond the 1 per cent level.
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
148
AMERICAN SOCIOLOGICALREVIEW garding middle class sensitivity to failure to realize rewards if the pattern is carried out uselessly. Fromm has contended, in other terminology, that disappointment and bitterness over the very slight results achieved from adherence to the DGP among lower middle class persons in Germany after World War I were of considerable force in helping Nazism to power. "If the savings of many years, for which one had sacrificed so many little pleasures, could be lost through no fault of one's own, what was the point in saving anyway?"16Here again we have relevance to the problem of order. It is an old and familiar contention that jobless intellectuals are a threat to any existing social system. In more general terms, to what extent does the failure of the DGP to "pay off" mean a revolutionary reaction, or perhaps merely a private neurotic reaction? And to what extent does the pattern have functional autonomy, so that while it may have had some initial partial motivation, at any rate, in considerations of expediency alone, it cannot be sloughed off and another way of life adopted at will even when it might be convenient to be able to do the latter? (3) In view of recent interest in class mobility, it may be suggested that one fruitful line of inquiry relates to the reaction of lower class individuals toward the incipient manifestation of tendencies toward the DGP on the part of others who are lower class. Would it be possible to develop a typology on pertinent lines? We may speculate that the following types of reactions would appear: (a) A reaction of lack of comprehension, conceivably rooting at times in lack of sympathy. "What's the point of this sort of behavior?" One goes out at fourteen or thereabouts to make a living, and no fuss about it. Anything else is not understandable. This sort of reaction may well require a rather firmly stabilized lower class tradition, not susceptible to the blandishments held out, for example, by higher education. (b) A "humble" reaction. "This education business (for example) isn't for the likes of me." Here we might well anticipate anger at others from the same class because of their "pretensions." (c) A derisive reaction,
16 Erich Fromm, Escape from Freedom, New York: Farrar and Rinehart, 1941, p. 214.
Students think that parents have saved money to give them a start in life. Class Index Self-Identification "Middle" "Working" Occupational Class 1 2 3 4 86 81 75 71 14 19 25 29 100 100 100 100 82 70 18 30 100 100 Yes No Total
* Figures in percentages. The chi square test shows differences are significant beyond the 1 per cent level for both class indices.
There is little point in elaborating the many qualifications on the significance in relation to the DGP of the above data. The professional sociologist can supply some of these for himself, but it appears obvious that the DGP is worthy of systematic research.
DISCUSSION
It is appropriateto add a few suggestions and questions relating to the theoretical significance of the pattern. (1) The pattern may turn out to have major significance for the "problem of order." Are there certain "advantages," certain socio-psychological "gains," that the so-called lower class has? It seems worthwhile to investigate the view that the lower class does have such advantages and reaps the "pleasures"of following impulse and not "deferring." Discounting the future, taking the cash and letting the credit go, disavowing major and worrisome self-disciplines, lower class persons may conceivably have a certain contentment that keeps them attached to an existing social order even when, from the point of view of other classes, they "live like animals."15 (2) The pattern carries implications re15 Perhaps the most fertile suggestions on this point have come from John Dollard, in his Caste
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions
INSTABILITIES IN STATUS on the line, "You're a sissy to study, to save, to defer." The above can only claim to be suggestions. They are deliberately neither sys-
149
tematic nor comprehensive. But they point to issues which are bound to come to the fore if sociologists address themselves to serious study of the DGP.
INSTABILITIES IN .STATUS: THE PROBLEM OF HIERARCHY IN THE COMMUNITY STUDY OF STATUS ARRANGEMENTS *
GREGORY P. STONE
and WILLIAM H.
FORM
tions have been raised against existing conceptions of social stratification.' While it is certainly to the credit of the Warner group that it has elicited such carefully considered and often trenchant discussions of social stratification, it is scarcely to the credit of the critics that so little research has been based upon their many well-taken criticisms. In their eagerness to disclose shortcomings in every phase of inquiry-methodological, conceptual, statistical, and ideological-the critics, with few exceptions, have only hinted at guide lines for future research.2 Beneath the controversy probably lie cer* This study is part of a larger researchproject sponsored by the Michigan State College Agricultural Experiment Station. The authors are indebted to Joel Smith for a critical reading of the manuscript. 1 The critical analyses in this period of stratification theory and researchwere undoubtedlyinitiated by C. Wright Mills' appraisal of W. Lloyd Warner and Paul S. Lunt, The Social Life of a ModernCommunity (New Haven: Yale University Press, 1942) in AmericanSociologicalReview, VII (April, 1942), pp. 263-2.71. 2Numbered among the exceptions would be Paul K. Hatt and Milton M. Gordon. See Paul K. Hatt, "Social Stratificationin the Mass Society," AmericanSociological Review, XV (April, 1950), pp. 216-222; and Milton M. Gordon,"A System of Social Class Analysis,"The Drew University Bulletin, XXXIX (August, 1951), pp. 3-19. The citation of these referencesdoes not imply that either contributionis acceptedhere without qualification. Specifically,the applicationof Hatt's paradigmfor the study of the mass society-presumably based upon Weber's distinctions among the social, economic, and political ordersof stratification-may be shown to violate the logic of Weber'sanalysis. As far as Gordon'sexcellentcontributionis concerned, the unqualifiedview of social stratificationas hierarchical in nature is a view to which this article takes exception.
"The Use of Class Concepts in Sociological Research," American Journal of Sociology, LIV (March, 1949); and, for a more empiricalcritique,
This content downloaded from 186.18.32.91 on Mon, 7 Oct 2013 22:09:52 PM All use subject to JSTOR Terms and Conditions