Subjective Elements in the Examination of Time Expenditures

Philip E. Hendrix, University of Michigan
ABSTRACT - Models of time allocation, following the cognitive approach to motivation, view individuals as striving to satisfy their needs by setting goals and choosing behavior that they believe will allow them to achieve these goals. Identification of these needs and their saliences is, of course, an essential task. Needs inferred from relatively objective characteristics of the individual (marital status, stage in family life cycle, etc.) have received more attention than the set of subjective needs (e.g., psychological - achievement, diversity, etc. - and social -affiliation, approval, etc.). Subjective characteristics of the individual, e.g., temporal orientation, have received even less attention. Reported here are some of the findings of a recent study in which elements of both sets of variables - objective and subjective - were included. A discussion of the usefulness of one of the subjective elements concludes the paper.
[ to cite ]:
Philip E. Hendrix (1980) ,"Subjective Elements in the Examination of Time Expenditures", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 437-441.

Advances in Consumer Research Volume 7, 1980     Pages 437-441


Philip E. Hendrix, University of Michigan

[The empirical results reported in this paper are drawn from data collected by the Institute for Social Research. The study, Time Use in Economic and Social Accounts, obtained four 24-hour time diaries from a national probability sample of some 1500 respondents and spouses between October, 1975, and September, 1976. The author would like to thank the principal investigators of the study for permitting access to the dataset.]

[The author is Visiting Assistant Professor of Marketing/ Faculty Associate in the Graduate School of Business Administration and Institute for Social Research, University of Michigan.]


Models of time allocation, following the cognitive approach to motivation, view individuals as striving to satisfy their needs by setting goals and choosing behavior that they believe will allow them to achieve these goals. Identification of these needs and their saliences is, of course, an essential task. Needs inferred from relatively objective characteristics of the individual (marital status, stage in family life cycle, etc.) have received more attention than the set of subjective needs (e.g., psychological - achievement, diversity, etc. - and social -affiliation, approval, etc.). Subjective characteristics of the individual, e.g., temporal orientation, have received even less attention. Reported here are some of the findings of a recent study in which elements of both sets of variables - objective and subjective - were included. A discussion of the usefulness of one of the subjective elements concludes the paper.


The relevance of the temporal dimension of consumer behavior, especially consumers' allocation of time, is well established (Becker, 1965; Schary, 1971; Jacoby et al., 1976; Nicosia and Mayer, 1976). Only recently, however, has the topic captured the attention of researchers. As one might suspect, important constructs are either lacking or inadequately specified in proposed models. In particular, the role of what may be termed subjective needs (including both psychological - achievement, diversity, etc. - and social - approval, affiliation, etc.) and subjective characteristics of the individual is not well developed. In the first section of this paper the premises of various theories of time allocation are presented, with the proponent's attention (or inattention) to subjective elements noted. The following section presents a subset of the results of a larger study of time allocation (Hendrix, 1978) in which several (of what I call) subjective [Subjective relative to the more objective characteristics of the individual (age, health, etc.) and his/her environment (marital status, stage in family life cycle, region of the country, type of dwelling, etc.)] elements were operationalized and their relationships with various time expenditures examined. One of these - the enjoyment corresponding to an activity - proved to be a strong explanator. The meaning of such a measure, the implications of alternative operationalizations, and potential uses are discussed in the final section.


The premise that time expenditures are intended to satisfy certain needs is central to most models of time allocation. For instance, in their study of time as a measure of household productivity, Walker and Woods (1976) note:

Household production changes over time within an individual family. The family is not a static entity but goes through stages of growth and contraction, with each stage requiring a different "mix", quantitatively and qualitatively, of goods and services to meet the needs of family members (p. 8) ... Thus, by knowing the composition of the household, it would be possible to predict how much time would have to be spent to produce the goods and services a family needed to function as a unit (p. 246).

Their hypothesis - that various family characteristics generate needs the satisfaction of which requires commensurate time expenditures - was generally supported by their analysis. Because of their limited objective - "... to develop a method for measuring goods and services that would be relatively convenient to use and for which data could be easily collected" (p. xiii) - Walker and Woods disregard psychological (subjective) characteristics of the individual. As they note (p. 8), "...such variations could only be assumed to average out over a sample of workers."

Arndt and Gr°nmo (1977) take a similar approach in positing "supply needs" (inferred from such data as number of children in family, stage in family life cycle, etc.) as an important determinant of the time devoted to shopping, though their data failed to support such a contention. They speculate that shopping may satisfy such "psycho-social" needs as diversion, self-gratification, and social interaction, factors which were not measured or tested in their analysis.

Chapin, (1974, p. 30) most explicitly recognizes the role of needs in his model of time allocation.

The energizing element in this model is the individual's ongoing drive to satisfy his or her wants, with the feedback of satisfaction-dissatisfaction levels of a previous experience in seeking to satisfy these wants regulating the intensity of the drive to engage in the behavior again.

In his initial analysis of the time spent in discretionary activities, Chapin examined the importance of two sets of factors: (1) "preconditioning factors," e.g., employment status, stage in life cycle, etc., which constrain the choices individuals have; and (2) such "predisposing factors" as felt need for status, achievement, and security. This latter set of variables provided a lower level of explanation than did the former, which he attributed to measurement and specification problems.

[Preconditioning factors] can be ascertained with little difficulty, but [predisposing factors] are less easily determined. Moreover, the full range of influences that prompt people to engage in an activity are likely to be less easily identified than those that constrain them from engaging in the activity (p. 210).

Chapin proposed a number of areas to which subsequent attention should be directed.

[It seems particularly important] to expand the scope of the predisposing element, bringing in other motivation factors and broadening this element to include enjoyment and thought-way factors [collective attitudes concerning particular activities] ... The use of a more encompassing range of felt needs appropriate to each generic class of activity would seem to merit exploration (pp. 210-11).

Perhaps the most comprehensive treatment of subjective elements in the examination of time expenditures is due to Morgan and his colleagues (1966). In their study of time spent in productive activities (paid work, housework, volunteer work, etc.), the impact of two sets of explanators was examined: "constraints and pressures" (including age, marital status, presence of children, etc.) and "motives and incentives" (achievement orientation, receptivity to change, closeness of family ties, mobility experience, planning and time horizon, etc.) Unfortunately, their precursory treatment of these subjective elements did not strongly influence subsequent work in the area. Based upon the assessment of current deficiencies in time allocation research by the leading theorists cited above, it seems that few advances have occurred. To continue to disregard or treat lightly subjective elements in the study of time allocation is, in my opinion, a mistake. Difficulties in measuring such elements may be more than offset by the gain in insight and level of understanding. The evidence reported in the next section supports this contention.


A theme common to the propositions of Morgan et al. (1966), Chapin (1974), and Arndt and Gronmo (1977) is that the time spent in a given activity is a function of two sets of determinants. One set of determinants (variously termed constraints and pressures, preconditioning factors, and antecedent conditions) stipulates some commensurate time expenditure in a given activity, e.g., an individual with small children will spend more time in preparing meals than the individual with no children, ceteris paribus. A conceptually distinct set of factors then accounts for variations about the time expenditure stipulated by the first set of determinants. Hendrix (1978) labeled the first set of determinants anterior conditions and the second set mediating factors. Postulated as members of the set of mediating factors were resources (income), time expenditures of others, energy expended, perceived role, temporal orientation, and enjoyment corresponding to the activity. These latter three variables may be considered subjective elements and thus will be elaborated.

Perceived role is defined as the extent to which an individual regards his or her own time input to an activity as an essential prerequisite to some outcome. In the conception of an activity as a process (Becker, 1965), the inputs of time and goods & services are combined in varying proportions to effect some outcome. One might expect the substitutability between an individual's time and various goods and services to be governed by such a role perception. To operationalize this construct, the relationships among three potential indicants were examined. It was hypothesized that an individual who perceived his or her own time input as essential would: (1) not desire more help from the spouse in caring for the children; (2) prefer having a good meal at home to eating out at a good restaurant; and (3) feel that a woman who works full time cannot have as warm and secure a relationship with her children as a mother who does not work. Since working with a secondary data set imposes severe constraints on the process of construct operationalization, it is not surprising that measurement was weak, e.g., the average inter-item correlation was .11. Nor was it surprising that the index comprised of these three elements did not contribute significantly to the explanation of time spent in any of four activities: meal preparation, meal cleanup, indoor cleaning, or procuring goods & services.

The ability of an individual to engage in activities in which the reinforcement is delayed was labeled temporal orientation. It was postulated that a present-oriented individual, one who requires immediate reinforcement, would fit the following profile: (1) feels that it is important to do the things he/she wants to do; (2) does not feel that it is important to do a job around the house right the first time, even if it takes more time; (3) waits until a job has to be done to do it; (4) usually buys something less expensive that won't last a long time; (5) lives from day to day; (6) does not do a lot of things which he does not like to do so that life will be better in the future; and (7) prefers that someone else do household chores so that he/she could do other things. Examination of the relationships among this set of variables via cluster and factor analysis suggested that variables (2), (3), (5), & (7) constituted the most homogeneous subset, though the average inter-item correlation (.154) was again low. An index was created using a weighted linear combination of these four variables. In analyses (using MCA) of the time spent in the four activities listed in the preceding paragraph as well as four additional ones - social entertainment, active leisure, viewing television, and reading - the following significant effects of temporal orientation were observed: present-oriented individuals spent more time (than future-oriented individuals) in social entertainment (p < .001), viewing television (p < .10), and reading (p < .025). Differences in the time spent by present and future oriented individuals in these three activities were substantial as well as significant, ranging (between the extreme categories of the index) from one to three hours per week.

The final subjective element incorporated in the study was the enjoyment corresponding to the particular activin. Respondents rated on a scale of zero (dislike a great deal) to ten (enjoy a great deal) some 22 activities. To improve the comparability of these ratings across respondents, the scores were standardized by individual, e.g., for each individual his or her ratings were used to compute a mean and standard deviation which were then used to convert the 22 "raw" scores to normalized scores.

Statistics computed from the normalized measures are interesting in and of themselves. Table 1 shows the ranks of the means and variances corresponding to the 22 activities. The most enjoyable (relatively speaking)activities included talking with and caring for children, talking with friends, taking children places, and going on trips and outings. The least enjoyable activities included cleaning house, grocery shopping, working in clubs and social organizations, and making repairs around the house. The most enjoyable activities were generally characterized by the smallest variances while the least enjoyable activities had the largest variances. The large variances which characterize the least enjoyable activities suggest that these activities are not universally unenjoyable. Thus, a useful avenue of research may be to identify segments of the total population for whom the reported measures of enjoyment corresponding to various activities differ.

The normalized scores were then bracketed for use as independent variables (along with other relevant variables) in the Multiple Classification Analysis of time expenditures in some seven activities. [In general, the continuous measures were bracketed into three or six categories so as to yield an approximately equal number of respondents in each category. In the analysis of time spent in meal preparation, meal cleanup and indoor cleaning, interactions between sex of the respondent and various independent variables necessitated disaggregation of the sample into male and female subsets. To insure that category sizes were sufficient to yield stable estimates, the number of categories was limited to three in these instances.] It was hypothesized that individuals who regarded an activity as relatively enjoyable would spend more time in the activity than individuals who regarded the activity as relatively unenjoyable. Results, shown in Table 2, confirm the importance of reported measures of enjoyment in the majority of activities examined. These findings are consistent with the results of two other studies.


Robinson (1977), using a measure of the satisfaction corresponding to activities, also observed the strongest relationships between the reported measures of affect and time spent in highly discretionary activities, prompting him (1977, p. 191) to conclude:

We suspect, therefore, that even daily routine evolves from a process in which individuals selectively find those activities that are psychologically rewarding and arrange their lives in such a way that participation in these activities can be scheduled more frequently.

Winter (1975), in an experimental setting, manipulated the attractiveness of alternative activities and examined the impact on the time which consumers spent in acquiring price information. Individuals faced with an attractive alternative use of their time spent less time acquiring price information than their counterparts faced with an unattractive alternative. The presumption is, of course, that subjects in both treatments allocated their time to the more attractive of the two alternatives.

To summarize, subjective elements have been incorporated to varying degrees in theoretical and empirical analyses of time expenditures. The inclusion of a wider range of these elements has been argued for by various theorists. Results presented here demonstrate the feasibility and usefulness of incorporating three of these subjective elements in analysis of time expenditures. To conclude, the role of one of these elements - enjoyment - is examined in more detail.



It was shown in the preceding section that a measure of affect - enjoyment - is capable of explaining variation in time expenditures. The role of such a measure should be Examined critically, however. Potential problems concern (1) semantics and (2) the level of understanding of behavior to which such a measure actually contributes.

Recall from above that various measures of affect have been used: Winter (1975) manipulated the "attractiveness" of alternative activities; Robinson (1977) measured the satisfaction corresponding to various activities, while the present study employed the term enjoyable. These terms are not interchangeable. The following discussion highlights the ambiguities which may arise. Reinforcement which an individual receives from engaging in an activity may be synchronous or asynchronous - simultaneous with or removed temporally from the expenditure of time, respectively. Using different labels for the same concepts, Juster (1978) suggests that activities may be characterized by process and/or outcome benefits. Of course, reinforcement, whether synchronous or asynchronous, may be negative as well as positive. A simple classification of activities based on the combinations of reinforcement is shown in Table 3.



Obviously, most preferred are cell 1 activities, characterized by synchronous positive reinforcement - enjoyable activities, if you will - and asynchronous positive reinforcement. Cell 3 is most likely a null set. If these two cells exhausted the possibilities, life would indeed be simple. Of course, they do not. Given the subjectivity of individual preferences as well as the options available for effecting certain outcomes, one must often trade off synchronous for asynchronous reinforcement and vice versa.

One might be forced to engage in an activity regarded as unenjoyable (in the individual's subjective estimation) in order to effect some desired outcome (cell 4) [Juster (1978) quite aptly calls these activities investment activities. The consumer foregoes present benefits (enjoyment) in order to obtain some future benefits, behavior which is, by definition, investment.] - examples might include such activities as home and auto maintenance, personal care, and education. Conversely, one might incur undesirable consequences in order to engage in an enjoyable activity (cell 2), e.g., overeating.

It should be apparent that activities in cells 1, 2, and 4 may be "satisfactory" uses of time - to each of them corresponds some type of positive reinforcement. Thus, the set of satisfactory activities (cells 1, 2, and 4) subsumes the set of enjoyable activities (cells 1 and 2) as well as those unenjoyable activities which result in desirable consequences. By specifying satisfactory activities, then, one fails to discriminate between activities which are enjoyable and activities which result in desirable outcomes. Information provided by a respondent which indicates that an activity is rewarding, acceptable, attractive, satisfactory, etc., does not enable the researcher to determine whether the activity is characterized by synchronous or asynchronous reinforcement.

Thus, researchers should take special care in selecting an affective term so as to avoid unintended interpretations. [Researchers in other fields have wrestled with this problem also. Jones and Pierce (1977, p. 298), in the social indicators literature, argue: "There is no reason ever to use the word 'satisfaction' or one of its should be suppressed in favor of less ambiguous terms descriptive of a person's affective state."] The least ambiguous of these terms appears to be enjoyment. At least, respondents seem to equate enjoyment with synchronous reinforcement.

In addition to the problem of semantics, another difficulty may arise in using measures of affect to explain time expenditures. In an earlier section, evidence was presented in support of the hypothesis that individuals seek to spend their time in what they regard as enjoyable activities. Thus, if one were to ask: "Why does John spend his time reading?" an empirically supported response may be "Because he enjoys reading." Such an explanation seems to require another question, namely, "Why does he enjoy reading?" As Aronson (1976, pp. 216-218) has noted:

Little understanding is gained by covering...behaviors with the blanket phrase 'reward', [for] in some way, all behaviors make sense, or feel good, or both, and can therefore be considered rewarding. But simply to label them as rewards tends to obscure the fact that there are important differences in kind among them...if our definition of what constitutes a reward is not clear [the concept] loses a good deal of its value.

Thus, we should recognize that the level of explanation due to measures of affect is superficial - that unless determinants of affect are specified, a less precise understanding of the behavior results.

Of course, the researcher must ultimately decide upon the level of explanation required. In some instances such "superficial" explanators as affect may be extremely useful. A current study, for example, is examining the relationship between the time spent in an activity and the corresponding measure of enjoyment to identify segments in which the hypothesized congruity is absent or relatively weak. The point is that researchers should not be deluded by such findings and neglect a more fundamental set of variables upon which both affect and behavior may be based.


Arndt, J. and Gr°nmo, S. (1977), "The Time Dimension of Shopping Behavior: Some Empirical Findings," Proceedings of the Association for Consumer Research Conference, 230-235.

Aronson, E., (1976), The Social Animal, 2nd ed., San Francisco: W. H. Freeman.

Becker, G. S. (1965), "A Theory of the Allocation of Time." Economic Journal, September 493-517.

Chaping, F. S., Jr. (1974), Human Activity Patterns in the City. New York: Wiley.

Hendrix, P. E. "The Allocation of Time by Consumers." Unpublished doctoral dissertation, The University of Michigan, 1978.

Jacoby, J., Szybillo, G. J., and Berning, C. K. (1976), "Time and Consumer Behavior: An Interdisciplinary Overview." Journal of Consumer Research, 2, 320-339.

Jones, M. B. and Pierce, J. M. (1977), "Time-Use Auditing: Approach to Validating Social Indicators." Social Indicators Research, 4, 289-315.

Juster, F. T. "The Investment of Time by U.S. Households: An Initial View." Paper presented to the American Psychological Association, 1978.

Morgan, J. N., Sirogeldin, I. A., and Baerwaldt, N. (1966), Productive Americans. Ann Arbor: Institute for Social Research, University of Michigan.

Nicosia, F. M. and Mayer, R.N. (1976), "Toward a Sociology of Consumption." Journal of Consumer Research, 65-75.

Robinson, J.P. (1977), How Americans Use Time. New York: Praeger Publishers.

Schary, P. B. (1971) "Consumption and the Problem of Time." Journal of Marketing, 35, April, 50-55.

Walker, K. and Woods, M. (1976), Time Use: A Measure of Household Production of Family Goods and Services. Washington, D.C.: American Home Economics Association.

Winter, F. W. (1975), "Laboratory Measurement of Response to Consumer Information." Journal of Marketing Research, 12, 390-401.