Demographics, Spending and Leisure: a Critique

Charles W. King, Purdue University
ABSTRACT - The research papers presented in the session, Demographics, Spending and Leisure, extended historical demographic analysis in the areas of social class and family life cycle. The research, however, used theoretical and methodological approaches developed over ten to twenty years ago. Demographic analysis in consumer research demands new multidimensional demographic constructs that reflect the contemporary cultural milieu.
[ to cite ]:
Charles W. King (1979) ,"Demographics, Spending and Leisure: a Critique", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 149-152.

Advances in Consumer Research Volume 6, 1979      Pages 149-152


Charles W. King, Purdue University


The research papers presented in the session, Demographics, Spending and Leisure, extended historical demographic analysis in the areas of social class and family life cycle. The research, however, used theoretical and methodological approaches developed over ten to twenty years ago. Demographic analysis in consumer research demands new multidimensional demographic constructs that reflect the contemporary cultural milieu.


The relationships between demographic variables and consumer behavior have been the focus of consumer research since the introduction of the analytical method in marketing. The relative effectiveness of demographic variables compared with other types of measures in explaining and predicting consumer behavior continues to be an unanswered question.

The three papers presented in the session, "Demographics, Spending and Leisure", relate traditional demographic variables to household expenditures by category and to leisure behavior. The papers represent a continuation of that historical demographic research tradition.

In reviewing the papers, significant issues of both a micro nature regarding the conclusions of the individual papers and of a macro nature regarding the entire concept of demographic analysis in consumer behavior have surfaced. Therefore, the critique of the papers has focused on:

1) the historical dimensionality of "demographic analysis" as reflected in the marketing literature;

2) a review and critique of the major conclusions and research issues presented in the papers;

3) the need for new demographic constructs in consumer research.


At the outset, a definition of terms is in order. What is "demographic analysis" as related to consumer behavior?

Demography is defined as "the statistical study of the characteristics of human populations, especially with reference to size and density, growth, distribution, migration, and vital statistics and the effect of all these on social and economic conditions" (Morris, 1971). Within the marketing context, demographic analysis involves "the study of market stratification of products or services in relation to such population characteristics as age, family size and type, sex, income, education and residence" (Ferber, et. al., 1964).

In the jargon of marketing, however, there is inconsistency in nomenclature regarding demographic and socioeconomic variables. Some researchers use the terms of demographic and socioeconomic variables interchangeably (Churchill, 1976). Other writers, however, draw clear differentiation between the two sets (Settle, Alreck, Belch, 1978). The pragmatic or theoretical implications of the differentiation (or lack of differentiation) are not clear in the literature. That may be reflective of the current state of maturation of demographic analysis as a measurement construct.

Demographic and socioeconomic characteristics, however, are routinely used as identifiers of key target market segments. The demographic and socioeconomic characteristics of consumer market segments have been referred to as "states of being" because they identify attributes or profiles of people (Buzzell, et. al., 1969).

Related to other types of variables, demographic and socioeconomic characteristics are often "enabling" variables that make possible various forms of consumer buying behavior. Though "enabling" in character, demographic and socioeconomic characteristics are usually not sufficient to guarantee a particular buying behavior.

A consumer, for example, may have adequate disposable personal income to participate in a particular leisure activity, e.g., skiing. The income level may be "enabling" -- necessary -- to afford the sport expense, but not sufficient to guarantee personal commitment/ involvement in that sport. "Intervening variables" in the cultural milieu and in life style preferences may channel the consumer buying behavior into other alternative leisure activities.

In terms of demographic analysis, the first step in analysis is to relate a specific, unidimensional variable, age or marital status, to some aspect of consumer behavior, e.g., expenditures or leisure activity. As an extension of the unidimensional variable approach, demographic variables have been operationally "configurated" or designed to be multidimensional. The new variables include combinations of standard demographic dimensions. Social class and family life cycle are two of the most widely used multidimensional demographic and socioeconomic variables measured.

Social class was first introduced as a socio-psychological concept in the United States over thirty years ago (Warner, 1949). Early pioneering work by the Chicago Tribune applied the notion in the marketing context in the 1950's (Martineau, 1957, 1958). Empirical definitions of social class typically involve the interactions of education, age, income and occasionally expenditure styles, e.g., expenditures on housing (Carmen, 1965).

The concept of family life style as a multidimensional demographic/socioeconomic variable was also developed in the marketing arena in the 1950's (Clark, 1955 and Wells and Gubar, 1966). The idea of family life cycle interrelates marital status, size of family, age of children, employment of family members, and residences of family members ("in the nest" or "out of the nest").

The basic postulate of family life cycle is that the process of family maturation over time can be organized around a series of discreet phases. Each phase has behavioral characteristics in terms of the social dynamics of the family unit. These dynamics ultimately impact on the family's consumer behavior in terms of both income availability and family needs and wants.

Interestingly, the concepts of social class and family life cycle as multidimensional demographic and socioeconomic variables are, in fact, first generation surrogate indicators of consumer life style segments. The validity of social class and family life cycle in consumer research is based on the crucial proposition that different social class segments or family life cycle stages are characterized by different behavioral patterns (life styles) that result in different consumer behavior. These variables preceded the development of life style and psychographics which emerged as a widely used research approach in the 1960's (Wells, 1974). Though some research has addressed this question, not enough interaction has developed between these closely related conceptual areas in the ensuing years (Reynolds and Wells, 1977).


An Overview

The research papers, in terms of a general overview evaluation, were conceptualized, empirically designed, and methodologically executed at the same level of research sophistication as the research traditions upon which they were built. Therein are both the strengths and weaknesses of the research contributions.

The papers universally represented extensions of the empirical data in the respective areas of social class and family life cycle research as applied to the analysis of consumer behavior. The papers did not, however, significantly extend the "state of the art" in the topic areas on either theoretical or methodological dimensions.

The Landon and Locander paper applied an established family life cycle paradigm with modifications to the analysis of leisure behavior research using a large sample survey research methodology. The conceptual approach, field measurement methodology, and analysis procedures of Arndt's research on family life cycle and household expenditure patterns used an established family life cycle measure with modifications. The field research survey was similar in overall design to statistical studies of consumer expenditures and construction/ analysis of the Consumer Price Index routinely conducted by the United States Bureau of Labor over the same period (Consumer Expenditure Survey Series, 1972-1973). Likewise, the Settle/ Alreck/ Belch research used routine demographic and socioeconomic measures and analysis in focusing on social class as related to standard types of consumer leisure activity.

Significant Contributions of the Research

Each of the research studies made its own unique contribution to its respective research tradition and associated literature.

Landon/Locander established the validity of the family life cycle concept as a useful tool in understanding leisure and recreation behavior. The authors note that the family life cycle concept has potential for explaining leisure behavior because "it matches needs with groups of people".

Arndt delivered a valuable contribution in the cross cultural data on Norwegian consumer spending patterns and relating them to the family life cycle construct. Arndt concluded that "size and composition of household expenditures are systematically related to stage in family life cycle", and summarized that these findings have implications for welfare programs and consumer policies. The Settle/ Alreck/ Belch research likewise related demographics and socioeconomic factors to leisure behavior. The researchers found that education was the best single socioeconomic determinant of leisure choices, contributing to the growing body of research on leisure behavior.

Significant Theoretical and Methodological Issues in the Research

In critiquing the research studies, significant theoretical and methodological issues were noted across the research projects. The research critiques have focused on four major dimensions for discussion: 1) literature review and theoretical rationale/"anchoring" of the research problems and designs; 2) construction of operational dependent variable measurements; 3) data collection and analytical procedures, and 4) interpretation of findings in terms of theory development or future research.

In the area of literature review and theoretical rationale/"anchoring" of the research problems and designs, all three papers could have presented more comprehensive reviews, research objectives and methodological rationale. The relevant theoretical frameworks, basic literature documentation and research issues were adequate and straight forward given the research objectives and the reporting medium.

The construction of the dependent variable measurements, however, lacked adaptive creativity and clarity of presentation. Admittedly, the measures were built upon historical research traditions of over twenty years and measurement comparability is essential for longitudinal comparison.

The issue is, however, that traditional dimensions of family life cycle and social class have potentially changed dramatically in the rapid socio/cultural transitions over the past decade. Those new dimensions were not explored or included in any of the operational measures used in the empirical research probing on family life style and social class as related to consumer behavior. Likewise, more detail on the specific measurements were needed in both the Landon/Locander and the Settle/Alreck/Belch papers.

On the dimension of data collection and analytical procedures, the Settle/ Alreck/ Belch data collection procedures involved an opportunity sample executed by 83 student interviewers conducting a class project. This raises the obvious problems of sample and interviewer methodology biases. Likewise, the use of a self-administered psychological test as an inducement for survey participation has serious implications in terms of bias, survey participant misinterpretation of the psychological test, and ethics of survey respondent manipulation.

On the analytical dimension, all of the reported analyses. suffered from a lack of analytical depth and/or extension. The basic empirical data content was rich. All of the research, however, could have benefited from use of a higher order of analytical sophistication.

Presentation of detailed tests of significance would have been useful in both the Landon/Locander and the Arndt research. The Settle /Alreck /Belch research could have employed multivariate statistical analysis in addition to its univariate approach.

On the dimension of interpretation of research findings in terms of theory development and future research, all of the research papers emphasized data presentation, tabular display, and "numbers reporting". The papers needed more qualitative analysis and interpretation of the data in terms of the respective research traditions and the "state of the art" of the topic areas upon which each project was built.

The Landon/Locander paper, for example, presented a very interesting data bank but concluded "FLC (family life cycle) offers a construct that could provide a rich independent variable to analyze present and anticipated recreational needs within the community. Likewise, FLC offers a means of segmenting markets into target groups whose research needs and subsequent behavior are relatively homogeneous." The researchers did not relate this data to the leisure research literature nor did they suggest issues for future research.

Arndt's expenditure pattern analysis could be viewed as merely a statistical report of Norwegian consumer expenditure patterns including a cursory qualitative summary of tabular results. By comparison, Arndt could have vastly enriched the research contribution by adding a cross cultural comparison of his data with similar, readily available published research conducted by the United States Bureau of Labor Statistics for the same year, 1972-1973. While the expenditure categories may not have been identically defined, there was enough comparability to support that comparative analysis and hypothesis generalization about cross cultural consumer expenditure dynamics.

By comparison, the Settle/Alreck/Belch research also failed to build any specific linkage between the research data and the research tradition with which the data related. On the other hand, these authors were very insightful in identifying and articulating the conceptual issues/challenges such as status inconsistency, the importance of various variables, e.g. education vs. income vs. occupational group, etc. in social class identification and/or determination. Additionally, the researchers presented a very valid discussion about the impact of cultural change on the concept and components of social class across the culture over time.


Analysis of the historical dimensionality of "demographic analysis" and review of the research papers in this session generate a central concluding thesis. There is a dramatic need for development of new multidimensional demographic constructs in consumer research.

The concepts of social class and family life cycle as multidimensional demographic measures applicable to consumer market segmentation were theoretically conceptualized and empirically operationalized over twenty years ago. Very little methodological research has focused on the validity of the concepts as historically "configurated" when applied in the contemporary era. The reported research used the concepts and their operational measures, with only minor modifications.

The cultural milieu has undergone traumatic social change over the past twenty years. That cultural change has impacted directly on the values of the social system and on the institution of the family.

Therefore, the central question is, are the historical measures of social class and family life cycle, in fact, applicable to the mass culture, circa 1978?

The Meaning of Social Class Under Social Change

The concept of social class, assuming Warner's original theoretical framework was valid, certainly deserves reassessment given the evolving values of the 1960's and 1970's. Social class is operationally defined within a specific societal context. Given the pluralistic character of the contemporary social milieu, can any single definition of social class be meaningfully applied across the social system?

To suggest there is no social class dynamic in the contemporary society would be naive. To suggest that the social class dynamic can be portrayed by one macro aggregate social class measure that is over twenty years old may also be naive.

Looking to the traditional elements of social class, e.g. education, income, occupation, are these still the relevant inputs? What are their respective weights in determining social what section of the social structure?

New social class measurements must be built upon assessment of the determinants of social class within particular societal situations. The current social class construct does not meet this challenge.

The Life Cycle and the Changing Family Structure

The concept of family life cycle may still be valid. Its operational measurement as reflected in the traditional nine stage continuum is not valid because it is not all inclusive. Over the past twenty years since the idea of family life cycle became popular, the institution of the family has experienced its own "future shock".

A variety of new family structures have evolved that are simply not identified in the current operational measurement of family life cycle. Because of the rising divorce rate, the one parent family has emerged as a structural entity that also flows through its own life cycle. The measurement of income, for example, becomes very complex within the divorced family community because of the confounding impact of alimony and child support payments on the disposable income of the receiver and on the payer. Likewise, the dynamics of child visitation cloud family membership measurement and significantly influence expenditure patterns.

Informal cohabitation is a type of family structure that has no convenient codification in the traditional family life cycle measures. That particular family arrangement, however, does influence consumer expenditure patterns particularly in housing and large ticket durable goods investments.

The working woman family has emerged over recent years as a unique type of family unit. Likewise, within the working woman families, an increasing share involve families with preschool children. In the context of family life cycle staging, the life style dynamics of this family segment are dramatically different from the classical, traditional profile of the "Full Nest, I (a)" stage as it has been defined in the operational measurements of family life cycle (Wells and Gubar, 1966). The implications for consumer behavior in the expenditure categories of child care and recreation are clear.

A Methodological Approach: The Convergence of Demographic Analysis and Life Style Research

The new demographic constructs can best be produced through a convergence of the two historical methodological traditions of demographic analysis and the maturing area of life style research. The multidimensional concepts of social class and family life cycle were both initially designed as surrogate life style descriptors prior to the development of the life style research tradition in the 1960-1970 era.

Social class and family life cycle, as demographic population descriptors, have been criticized, not because of their conceptual validity, but because of their operational measurement weaknesses. Nor are these two dimensions the only potential multidimensional demographic measurements that might be informative to consumer researchers.

A complete review of demographic analysis is needed to identify relevant target consumer segments endemic to the contemporary culture. Life style dimensions must be explicitly combined with traditional unidimensional demographic variables such as age, income, education, marital status, employment, etc., to produce richer, more accurate and more relevant descriptions of the "status of being" of the evolving consumer market structure.


Robert D. Buzzell, Donald F. Cox, and Rex V. Brown, Marketing Research and Information Systems: Text and Cases (New York: McGraw Hill, 1969)

James A. Carman, The Application of Social Class in Market Segmentation (Berkeley: Institute of Business and Economic Research, University of California, 1965)

Gilbert A. Churchill, Jr., Marketing Research (Hinsdale, Illinois: The Dryden Press, 1976).

Lincoln H. Clark, ed. Consumer Behavior, Vol. II. The Life Cycle and Consumer Behavior (New York: New York University Press, 1955)

Consumer Expenditure Survey Series: Interview Survey, 1972-1973, Report 455-3 (Bureau of Labor Statistics, Department of Labor)

Robert Ferber, Donald F. Blankertz, Sidney Hollander, Jr., Marketing Research (The Ronald Press: New York, 1964)

P. Martineau, Motivation in Advertising (New York: McGraw-Hill Book Company, 1957)

P. Martineau, "Social Classes and Spending Behavior," Journal of Marketing, 23(October, 1958), 121-130.

William Morris, ed. The American Heritage Dictionary of the English Language (New York: Houghton Mifflin Company, 1971)

Fred D. Reynolds and William D. Wells, Consumer Behavior (New York: McGraw-Hill, 1977)

W. L. Warner, Social Class in America - The Evaluation of Status (Chicago: Science Research Associates, Inc., 1949)

William D. Wells, ed. Life Style and Psychographics (Chicago: The American Marketing Association, 1974)

William D. Wells and George Gubar, "Life Cycle Concept in Marketing Research," Journal of Marketing Research, (1966), 355-363.