Family Life Cycle As a Determinant of Size and Composition of Household Expenditures

ABSTRACT - This study is concerned with the family life cycle as a predictor of spending patterns. In a national probability sample of Norwegian households stage in family life cycle was related both to size of expenditures and their allocation to product and service groups and to gifts given.


Johan Arndt (1979) ,"Family Life Cycle As a Determinant of Size and Composition of Household Expenditures", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 128-132.

Advances in Consumer Research Volume 6, 1979      Pages 128-132


Johan Arndt, Norwegian School of Economics and Business Administration


This study is concerned with the family life cycle as a predictor of spending patterns. In a national probability sample of Norwegian households stage in family life cycle was related both to size of expenditures and their allocation to product and service groups and to gifts given.


In consumer behavior the family life cycle notion is one of the most over-quoted and under-researched concepts. The family life cycle is an idealized construct referring to the fact that the typical family passes through a fixed sequence of important stages in its life. These stages affect the economic behavior of the family.

Most consumer behavior textbooks, such as Engel, Kollat, and Blackwell (1973, pp. 193-196), Block and Roering (1976, pp. 138-142), and Berkman and Gilson (1978, pp. 204-206), routinely discuss the life cycle concept and its implications. But the studies quoted tend to be dated ever-greens. In fact, some of the most important contributions on the subject matter were presented at a conference at Ann Arbor, Michigan a quarter of a century ago (Clark, 1955). For instance, in one major paper, Lansing and Morgan (1955) showed how income, expenditures on durable goods, assets and debts, and subjective feelings about the financial situation depended on stage in the life cycle. Other relevant papers were delivered by Barton (1955), Fisher (1955), and Miller (1955). These and other studies such as Lansing and Kish (1957) were reviewed a decade ago by Wells and Gubar (1966).

However, after this promising start, there has been a long hiatus in which few memorable new studies have emerged. An exception is the mainly unsuccessful attempt by Rich & Jain (1968) to explain variations in women's shopping by stage in life cycle. A partial explanation of the neglect of the family life cycle is the dominating interest in studies having the individual consumer (rather than the household) as the unit of analysis. Another reason may be the shift away from single constructs such as the family life cycle to more comprehensive theories.

The purpose of this article is to contribute to a recycling of the family life cycle concept. As appears from the Wells and Gubar (1966) review, past contributions have mainly treated the life cycle as an independent variable to account for differences in stocks (product ownership and debts) and flows (incomes and expenditures of different kinds).

Past studies have mostly been concerned with isolated, fragmentized aspects of economic behavior. This study attempts to make a more "holistic" approach in the sense that it addresses the total amount of household expenditures and its allocation to products and services bought, taxes, social insurances, and fees paid, and gifts given. A limitation of this study (and its predecessors) is that it is essentially a distributive study focusing on behavioral outcomes of decision processes rather than the process leading to the outcomes (Lazarsfeld, 1959).

Before describing the method and presenting the results of the study, a few comments on the theoretical importance of the family life cycle concept are in order.


The family life cycle notion refers to the important transitions or critical passages in the life of the typical nuclear family. Inside the family, the social roles of its members differ over the life cycle. Moreover, externally, the social position of the family in relation to society is partially determined by stage in the life cycle.

In the pre-industrial society, the family (mainly the extended family) was not only the appropriate unit of analysis for consumption activities, also most of the production functions were carried out in the context of the family. In modern society these two functions have been split, leaving only consumption activities as the domain of households or families. Even in this area there are indications that the role of the household is decreasing (dining out, prepared foods, etc.).

To the student of consumer behavior, family life cycle is of some interest as it is related both to the economic resources (purchasing power) of the family and to the needs and wants manifested in the spending of the money.

Lansing and Kish (1957) and Lansing and Morgan (1955) found that over the life cycle the total incomes of the spending units increased to a peak and fell off in the last stages. However, there was a definite dip for young married couples with young children due to the withdrawal of wives from the labor force.

Family life cycle affects the composition of consumer expenditures in two ways. First, there is the income effect referring to the fact that products and services differ in income elasticity (as pointed out, incomes vary over the life cycle). Necessities such as food, clothes, and minimum housing expenses are fairly inelastic with regard to income. However, income increases result in disproportionate increases for "luxury" or "discretionary'' items such as cars, expensive appliances, recreation items and gifts given. Second, stage in family life cycle may be used as a proxy or surrogate indicator of needs and wants. For instance, young couples with children have a higher need for food and clothing, and may be expected to spend a disproportionate part of their incomes on these categories.

The simplified framework underlying the discussion above is summarized in Table 1.

A recurring problem with family life cycle research is the difficulty of developing a meaningful categorization of the life cycle, As observed by Wells and Gubar (1966, p. 360), not two investigators have yet agreed on which categories to use. Needless to say, this has hindered comparisons and the development of a cumulative research tradition. In brief, the trade-off problem is as follows: There will be too large within-class variation (and too little between-class variation) if the taxonomy contains a small number of classes. On the other hand, a large number of categories may mean that many classes may be unpopulated unless large-scale surveys are involved.



In the study to be presented here, the following six-class scheme was used: (1) Bachelor, (2) Newly married, (3) Full nest I, (4) Full nest II, (5) Empty nest, and (6) Solitary survivors. The criteria are shown in Table 2 below. The categorization which partly was determined by the secondary nature of the data base, differs somewhat from the nine-category scheme recommended by Wells and Gubar (1966). The differences are as follows: Wells and Gubar (1966) recommend distinguishing (1) between Full nest I families with youngest child under six and others, (2) between Empty nest families with head of household in labor force and others, and (3) between Solitary survivors in labor force and those who are retired.

Table 2 below summarizes the theoretical expectations for this study based on the general discussion above, the specific findings reviewed by Wells and Gubar (1966), and some common-sense. The central idea is that economic resources differ by stage in family life cycle, and that each stage has an appropriate "standard package" of goods and services.


Ideally, a longitudinal (panel) study would be the appropriate research strategy for examining consumer behavior over the family life cycle. Such an approach would make it possible to observe changes in income and demand composition when they occur. Unfortunately, such an approach is infeasible in practice because of the high costs involved, and, more important, the long time perspective required - literally the whole professional life of a researcher.

Hence, this study like preceding studies (Wells and Gubar 1966) had to resort to a less than ideal cross-sectional design. Such an approach permits making inferences about what happens at the various life cycle stages by comparing households in different stages.

The source of the data was the Survey of Consumer Expenditure 1973 conducted by the Central Bureau of Statistics of Norway. An original purpose of this study was to update the weights used to calculate the consumer price index. The unit of analysis in this survey was the household, defined as all persons having a common dwelling and sharing at least one meal per day, excluding all institutional households such as hospitals.




The sample consisted of 4,707 households selected by a stratified, self-weighting two-stage probability sampling method, ensuring a nationally representative sample of households.

Each of the households living at the 4,707 drawn addresses were contacted by interviewers and asked to participate in a personal interview (collecting information about name of household members, date and year of birth, household and marital status, and occupation and occupational status). The second source of data was detailed account books in which respondents (in return for a small honorarium) were requested to enter all expenditures over a two-week period on products and services bought, payments for taxes, social insurances, and fees, and the value of gifts given. Consumption of own-produced products (milk, butter, potatoes, fresh meat, etc.) were also recorded. Third, when the account books were collected, a second personal interview gathered information about expenditures on products with low purchase frequencies (including certain expenses for dwelling, car purchase, and purchase of special appliances) and information about product ownership. As size and composition of household expenditures vary with season, data had to be collected over a whole year. For instance, 1/26 of the households kept accounts between January 1 and 14, 1973, another 1/26 between January 15 and 28 and so on.

In all, 3,363 households (or 71 per cent) completed the account books and participated in both interviews. The most important reasons for non-response were refusals to participate and absence from the address.


Stage in Family Life Cycle. The 3,363 households were classified by stage in the life cycle on the basis of the demographic information from the introductory interview. The distribution is shown in Table 3. A reason for the large relative size of the "Other Households" class was the fact that the upper age limit for dependent children was 16 years (set by the Central Bureau of Statistics).



The account books were the primary source of expenditure data. The record-keeping was very detailed. For instance, distinction was made between five different varieties of bread. Here we will be concerned with (aggregated) main groups of consumption expenditures (CE). It was here distinguished among the following nine expenditure groups (EG): (1) Food, (2,) Beverages and tobacco, (3) Clothing and footwear, (4) Housing (rent, fuel, and power), (5) Furniture and household equipment (appliances), (6) Medical care, (7) Transport (car, gasoline, public transportation, etc.), (8) Recreation and education (television sets, boats, movies, theaters, books, etc.), and (9) Other goods and services (restaurants, hotels, packaged tours, personal care articles, etc.). For each household, the annual expenditures in each of the nine groups were projected.

Second, all expenditures on taxes, social insurances, and fees (TSF) were also recorded in the account books, along with the monetary amount of gifts given (G).

The total household expenditures (TE) were accordingly composed as follows:

TE = CE + TSF + G  (1)



In this study total household expenditures (TE) may be treated as an indicator of household income. It should be noted, however, that savings and investments in real estate are not, included in TE. However, TE is believed to be strongly correlated with household incomes.


The size and composition of consumption expenditures by stage in family life cycle are shown in Table 4. As seen in Table 4, the average consumption expenditures (CE) per household were Norwegian Kroner 36,832 in 1973 (a Norwegian Kroner is about U.S. $20). Compared with the findings from the corresponding survey in 1967, this indicates an increase in expenditures of 75 per cent. During this period the consumer price index rose by 44 per cent.

The most important expenditure group was Food having a budget share of 24 per cent. This is a substantial relative decrease from 1967 when the food share was 30 per cent. On the other hand, the shares of Recreation and education (9.7 per cent) and Transport (19.5 per cent) increased significantly from 1967. Such patterns as found here are common in societies in fairly rapid economic growth.

Below, first the results relating to size of expenditures will be discussed. Next, findings with regard to the composition of consumption expenditures and gift-giving will be presented.

Size of Expenditures

Total household expenditures (TE) are shown in Table 5 (4th row). Like in the earlier American studies (Lansing and Morgan, 1955 and Lansing and Kish, 1957), the lowest household incomes were at the extreme ends of the scale, among Bachelors and Solitary survivors. The relative levels of the lowest to the highest incomes were also similar to the U.S. patterns. Moreover, in support of the theoretical expectations (and earlier empirical findings), there was a dip in the expenditures curve for the Full nest I stage. The highest total expenditures were found for Newly marrieds. Also Full nest II families had high expenditures.





As seen in Table 5 (second row), taxes, social insurances, and fees (TSF) made up about one-quarter of total household expenditures. But this proportion varied with stage. Hence, in spite of having 5 per cent lower total expenditures (TE) than Newly marrieds, Full nest I households had approximately the same consumption expenditures (CE) because of lower taxes. However, viewed differently, since there are more mouths to feed in this stage there were, of course, significantly smaller consumption expenditures per family member. Hence, in spite of the tax advantage for this group, the standard of living may have been lower than in the previous stage.

Composition of Consumption Expenditures

As predicted, Bachelors were found to spend disproportionately much on "luxuries" such as restaurants and entertainment (Other Roods and services in Table 4), cars (Transport), and Recreation an<] education, while relatively little on Food, Housing and Furniture and Household equipment.

In contrast, Newly marrieds though having almost double the income spent relatively more on Food, Housing, and Furniture and household equipment. On the other hand, they saved relatively on Clothing and footwear, Recreation and education, and Other goods and services.

When the first child arrives and the household is moved on to the Full nest I stage, the result was as expected higher shares of consumption expenditures for Food, Clothing and footwear, and Housing compensated by savings on Transport, and particularly Other goods and services (apparently curtailing eating out substantially).

The results conform with the expectations also for the fourth stage. Full nest II households reported spending relatively much on Food and Clothing and footwear. However, the predicted high share for Furniture and household equipment did not show up.

The most dramatic change from stage four to five related to total expenditures which shrunk by about 40 per cent. As seen in Table 4, the Empty nest group exhibited many of the patterns predicted for the last stage: high shares for Food and Medical care, and low shares for Recreation and education and Other goods and services. Another surprising finding was the high share for Beverages and tobacco. This group contained a fairly high share of economically inactive persons (households) and also persons with relatively low education.

For the Solitary survivors group the results were as predicted. This group had the lowest household income combined with the highest share of Housing costs, what suggests severe problems in making ends meet.


The findings relating to gift-giving are shown in Table 5 (third and fifth rows). It appears that in absolute terms, expenditures on gifts were surprisingly constant over the life cycle. However, taken as a proportion of total expenditures it appears that the Newly marrieds and the Full nest I and II households as expected had least to spare. The by far most generous group were Solitary survivors, who, as earlier pointed out, seemed to be in the least favorable economic situation. This may be a parallel to the biblical widow who was generous in spite of her poverty. Another explanation is that this group may also contain persons who feel that most of their own needs are met and, therefore, have surplus money.


In general, the results of the study tended to support the notion that size and composition of household expenditures are systematically related to stage in family life cycle.

The findings of this and earlier studies have obvious implications for welfare programs and consumer policies. For instance, public consumer programs are likely to have different impact depending on stage in life cycle of the citizens involved. Since consumer resources and needs vary over the life cycle, consumer information programs should be tailored to fit these differences.

Marketing managers may use family life cycle as one out of several inputs in forecasts. Demand for different product and service categories may be estimated from knowledge of the relationship between demand and stage in life cycle and the predicted number of households in the various stages. Second, family life cycle may be used as a criterion for segmenting markets, identifying the most likely buyers.

The nuclear family is presently under pressure. In the future, it may be increasingly difficult to classify consumers by stage in family life cycle because of informal cohabitation arrangements, and a growing number of divorces and remarriages. For the time being, however, the concept seems to be meaningful in consumer research.


S. G. Barton, "The Life Cycle and Buying Patterns," in Consumer Behavior, Vol. 11. The Life Cycle and Consumer Behavior, ed. Lincoln H. Clark, New York: New York University Press, 1955, 53-57.

Carl E. Block and Kenneth J. Roering, Essentials of Consumer Behavior, Hinsdale, Illinois: The Dryden Press 1976.

Harold W. Berkman and Christopher C. Gilson, Consumer Behavior, (Encino, California: Dickenson Publishing Company, 1978).

Lincoln H. Clark, ed. Consumer Behavior, Vol. 11. The Life Cycle and Consumer Behavior, (New York: New York University Press, 1955).

James F. Engel, David T. Kollat, and Roger D. Blackwell, Consumer Behavior, 2nd Edition, (New York: Holt, Rinehart and Winston, 1973).

Janet A. Fisher, "Family Life Cycle Analysis in Research on Consumer Behavior," in Consumer Behavior, Vol. 11. The Life Cycle and Consumer Behavior, ed. Lincoln H. Clark, (New York: New York University Press, 1955), 28-35.

John B. Lansing and Leslie Kish, "Family Life Cycle as an Independent Variable," American Sociological Review, 22(1957), 512-9.

John B. Lansing and James N. Morgan, "Consumer Finances Over The Life Cycle," in Consumer Behavior, Vol. 11, The Life Cycle and Consumer Behavior, ed. Lincoln H. Clark, (New York: New York University Press, 1955) 36-51.

Paul F. Lazarsfeld, "Sociological Reflections on Business: Consumers and Managers," in Social Science Research on Business: Product and Potential, eds. Robert A. Dahl, Mason Haire, and Paul F. Lazarsfeld, (New York: Columbia University Press , 1959) 99-155.

Donald L. Miller, "The Life Cycle and the Impact of Advertising,'' in Consumer Behavior, Vol. 11. The Life Cycle and Consumer Behavior, ed. Lincoln H. Clark, (New York: New York University Press 1975) 61-5.

Stuart U. Rich and Subhash C. Jain, "Social Class and Life Cycle as Predictors of Shopping Behavior,", Journal of Marketing Research, 5(1968) 41-9.

William D. Wells and George Gubar, "Life Cycle Concept in Marketing Research," Journal of Marketing Research, 3(1966) 355-63.



Johan Arndt, Norwegian School of Economics and Business Administration


NA - Advances in Consumer Research Volume 06 | 1979

Share Proceeding

Featured papers

See More


J1. The Effect of Identity Abstractness on Information Processing Styles

Woojin Choi, University of Seoul
Min Jung Kim, Manhattan College
HyukJin Kwon, University of Seoul
Jiyun Kang, Texas State University

Read More


Is All That Glitters Gold? The Effect of Product Surface Glossiness on Consumer Judgments

Jiaqi SONG, Hong Kong Polytechic University
Yuwei Jiang, Hong Kong Polytechic University
Gerald J. Gorn, Hong Kong Polytechic University

Read More


Trapped in the rabbit hole: Life history strategies modulate the impact of mild stress on hedonic consumption

Bob Fennis, University of Groningen, The Netherlands
Justina Gineikiene, ISM University of Management and Economics
Dovile Barauskaite, ISM University of Management and Economics
Guido M. van Koningsbruggen, Vrije Universiteit Amsterdam

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.