A time budget is a log or diary of the sequence and duration of activities engaged in by an individual over a specified period, most typically the 24-hour day. Time-budget research involves the collection of numerous such protocols from members of a population to analyze main trends and subgroup differences in the allocation of time.
While many social inquiries are concerned with the amount of time spent on a particular kind of activity (the journey to work or time given to television watching), the term “time budget” is generally reserved for an exhaustive accounting of a slice of time in individual experience, whatever the component activities happen to be. Such time budgets naturally yield information on the time consumed by any of the wide range of activities in which man engages. Yet at the same stroke, this information is given a context, since the protocols pose the question of, for example, what else it is that people who spend little time on television do with the time thereby saved or of where the time comes from for those who devote enormous quantities to it. One of the special assets of time-budget research is the immutable fact that no human being, however rich, poor, wise, or foolish, can dispose of more time than any other within the same period. Hence, variations in time allocations from person to person must depend on “trading off” time from some activities toward others.
Indeed, the phrase “time budget” has arisen because time, like money, is a resource that is continually being allocated by the individual, although with varying degrees of consciousness and short-term discretion. Like money, time is thought of as being spent, saved, invested, or wasted. It is presumed that analysis of the structure of time allocation gives behavioral evidence of a peculiarly “hard” kind concerning individual preferences and values, especially in the more optional forms of time use.
The time budget as a research tool has been employed in such areas as the study of life styles, the sociology of leisure, and research on aging and other aspects of the life cycle. Collection of systematic time-budget data is a relatively recent activity. Such data also provide material for curiosity concerning long-term secular trends in the “quality of life,” such as the division of time between work and play (de Grazia 1962) and the role of system-level work inputs in economic development.
The systematic collection of family financial budgets represents one of the earliest forms of empirical research at a microsociological level, with examples dating back to seventeenth-century Europe and coming to flower in the celebrated work of Frederic Le Play in the nineteenth century. The extension of the paradigm to time expenditure arrived somewhat more slowly, however. Friedrich Engels’ The Situation of the Working Class in England, first published in 1845, contains a wide variety of information on the daily round of industrial workers, with some quantitative estimates of time expenditures. Perhaps the first careful and exhaustive time budgets covering at least the period of work were represented in the time-and-motion studies of Frederick Taylor in his attempt, around the turn of the century, to institute “scientific management.”
The first large-scale study of exhaustive 24-hour time budgets was carried out in 1924 on Moscow workers by Stanislav G. Strumilin, as part of the postrevolutionary drive toward more rational economic planning [seeleisure]. In the West, most comparable work during the same period was focused on leisure, ignoring the portions of the day that were denned in some sense as nonleisure. However, near the end of the 1920s the Bureau of Home Economics of the U.S. Department of Agriculture conducted five studies of time use among farm women that involved full time budgets in the now customary sense of the phrase. By the 1930s more studies were beginning to appear. Lundberg and his associates (1934) based their intensive examination of suburban life in West-chester County in part on nearly 5,000 time budgets drawn from about 2,500 respondents. Sorokin and Berger’s Time-budgets of Human Behavior (1939) attempted to use such materials to investigate the structure of human motives and in so doing provided one of the more germinal surveys of the possibilities of the time budget as a research tool.
Along with the rapid expansion of survey research after World War n, the generation of time-budget data has shown an enormous increase, although this trend is less prominent in the United States than in most other industrialized nations. Part of the impetus for this research has come from commercial and governmental enterprises concerned with the timing of activities in their mass populations. Perhaps the largest single time-budget survey, involving a sample of 170,000 persons, was carried out in Japan in 1960-1961 to aid in the scheduling of radio and television programming (Nakanishi 1963). One of the few nationwide time-budget surveys in the United States was conducted in 1954 for similar reasons by the Mutual Broadcasting Company, and comparable, although smaller, surveys for the mass media have been done in most countries of western Europe. Another impetus to such work has come from the economic-planning needs of governments. After a period in which social research of all kinds was neglected, the tradition of time-budget surveys established by Strumilin in the Soviet Union was resumed on a massive scale in the late 1950s. In addition to thorough methodological work, the monitoring of time budgets has become sufficiently institutionalized in the Soviet Union for national scientific conferences to have been held for the purpose of establishing uniform codes and standards, and several hundred thousand man-days’ worth of time-budget material has been accumulated (Szalai 1966). In a similar vein, time-budget data were gathered in Hungary in 1963 by the Central Statistical Office, as part of the official microcensus. Quite generally speaking, time-budget work represents one of the most vigorous traditions of social research in eastern Europe. Western social scientists have used the technique much more sparingly, although often with an eye toward broadening its utility.
The most ambitious effort at a rigorously comparative cross-national time-budget study was launched in 1964 under the sponsorship of UNESCO and the International Social Science Council. Directed by Alexander Szalai, research institutes in Bulgaria, Belgium, Czechoslovakia, France, East and West Germany, Hungary, Peru, Poland, the Soviet Union, the United States, and Yugoslavia had by 1966 successfully collected time-budget information on probability samples of urban populations in their respective countries, following standardized methods of interviewing and coding. Extensions of the study design to Greece, Cuba, and Canada were also projected (United Nations …1966).
The first decision to be faced by the student of time use has to do with choice of time scale for his budget collection. Some time budgets are organized to cover time spent at the workplace; others span only “leisure hours.” At the opposite extreme, it is possible to imagine time budgets covering an individual’s life cycle, yielding information of totally different significance, although nothing but the most sketchy efforts of this kind exist at present [seelife cycle]. There are grounds for arguing that estimates of productive work time should be based on time budgets covering annual cycles, thereby including vacations and a reasonable sampling of illnesses.
In point of fact, however, the vast majority of time-budget research is geared to the 24 hours of a single day in the life of a respondent. By the gathering of enough time diaries, dispersed over weekdays and weekends as well as across the seasons of a year, from various individuals of differing ages, a synthetic picture of weekly, annual, and life-cycle variations can be organized from single-day protocols. In those rare instances where individuals have kept diaries for much longer than a single day or two, results have been reassuring. For example, Lundberg and his associates (1934, appendix) reported, among other things, that time diaries showed high stability in activity allocations for respondents maintaining them for longer periods.
At first glance, the collection of 24-hour time budgets would seem to be a relatively simple and straightforward task. Indeed, the time continuum itself offers a quantitative variable of a purity and precision rarely found in social research, and one readily understood by almost any respondent in industrialized societies. Yet closer familiarity makes clear that the gathering of even day-length time-budget information poses an oppressive array of complexities. The method shares all of the common problems of reliable population sampling and data elicitation and adds many unique burdens as well.
Though time may be a “clean” and “hard” variable, activity units are not. How finely should we slice in establishing an array of activities? Chronicles of “a day’s activities” might be so grossly characterized as to show but six or seven activities; or again, they might reflect such a fine grain of movement and action as to number two hundred or more for the same day. Some activities, such as sleeping, are rather clearly demarcated. Others are punctuated by major spatial displacements ("drove to the drugstore"). For many activities, however, lines of demarcation blur, and the common guide-posts become deceptive. Is “work at the office” simply to be left as an undifferentiated block of time, or is it to be thinly subdivided? “Shopping” often involves a sequence of spatial displacements. Is it one activity by itself or one for each shop visited, each purchase made, or what? And more pressing still, how is the researcher to handle simultaneous activities, such as cooking dinner while listening to the radio and keeping an eye on the children?
Increasingly, investigators have tried to depend on the respondent himself to define what an activity is by the way he chooses to record the events of his day. Thus, one method of data collection— asking the subject to check activities from a pre-designated list of possibilities—tends to be avoided as imposing too much outside structure on spontaneous perceptions. Nevertheless, when “raw,” or volunteered, descriptions are accepted, some of the burden returns to the investigator in the form of decisions as to how the concrete acts are to be coded into more general classes. And whether the respondent is asked to “precode” or the coding is accomplished later, there is an intrusion on reality which can affect results in vital ways.
There are undoubtedly real differences in activity levels between persons, from the bedridden invalid to the harassed professional, so that the sheer number of activities per day should in fact vary. But some of the recorded interpersonal variation can be due to different perceptions of the task instructions, and the whole texture of findings can be displaced by nearly arbitrary decisions on the part of the investigator in providing task instructions or in fixing the frequency with which activities should be recorded (for instance, should it be every five minutes, every hour, or four times a day?). In general, the more activities are recorded or coded in gross fashion, the more alike people of different social statuses appear to be.
In order to keep the time chronicle totaling a tidy 24 hours, researchers have often decided to discard the “less important” of any activities carried on simultaneously, although any such decisions lead immediately to new problems concerning the definition of importance. More recently, however, secondary and tertiary activities are being maintained as separate parts of the record, and intriguing work on the incidence and structure of such simultaneous activities is under way (Guilbert et al. 1965).
Similarly, while some analytic grouping (that is, coding) of activities is always required, the trend is toward increasingly detailed activity codes that hew more closely to concrete reality, at least in the first stage of analysis. Subsequently, such codes permit greater flexibility in regrouping elements for differing analytic purposes. The standardized code for time-budget research in the Soviet Union involves 99 categories; the UNESCO multination project proceeds at the same level of detail.
Broadly speaking, there are three main methods of eliciting 24-hour time budgets from respondents. In the diary method, the respondent simply writes down his own log of activities and their durations, more or less synchronously, on the basis of instructions and forms provided. Drawbacks of such a straightforward approach include the fact that subjects show considerable variation in the diligence and care with which such diaries are maintained, and the protocols are thus very uneven in quality. Moreover, functionally illiterate persons can be frightened by such a task, and a distressingly large proportion of them refuse to cooperate. A second method, the “yesterday interview,” requires the subject to reconstruct his previous day’s sequence of activities orally, under probing from an interviewer, and to estimate times spent on each until the full period of 24 hours has been accounted for. The limitations of this approach are obvious: people forget. Experimental evidence as to differences between perceived time and physical time abound, and many details are completely lost to memory at only a day’s remove. A third method involves observation and recording of the individual’s activities by a second party. However, such an approach tends to have low feasibility for both cost and privacy reasons, and it has become apparent that the outside observer has much more difficulty deducing meaningful “activity units” than does the actor himself.
While most research in the past has depended on the unsupported diary method, the current trend is toward overcoming the liabilities of each method by a combined approach. Thus, for example, a subject may be asked to maintain a diary for 24 hours in the standard manner, but the log itself may be used as merely a refresher or initial skeleton for more detailed discussion in a “yesterday interview.” Even a sketchy diary will put stringent limits on the respondent’s imagination as far as time durations are concerned, yet the interviewing can serve to bring the account of activities on the part of a poorly educated subject to a level of detail that a well-educated subject may be able to capture by himself.
Whenever a diary or an outside observer is involved, a burden is placed on the respondent which considerably exceeds that of a standard sample-survey interview. Therefore, refusal rates mount from the normal insignificant levels to heights that daunt the most experienced researcher. An embarrassing proportion of the time-budget literature from the past has been generated from captive audiences like schoolchildren, or suffers refusal rates so high as to call the data into question at the outset, or has paid no attention whatever to representative sampling. One of the time-budget studies which is most challenging at a conceptual level (Sorokin & Berger 1939) is based on about 5 per cent returns of diary forms distributed to workers on relief and white-collar unemployed in the Boston area during the depression of the 1930s. More generally, it is obvious that potential respondents who refuse to participate because they are “too busy” strike to the very heart of the purposes of the study. Although it has been shown possible in recent years to approach a probability sample of a cross-section population with the 24-hour time-budget task and keep refusals within tolerable limits (20 per cent, for example), such success depends on elaborately planned approaches and inducements.
The variety of complexities in time-budget methodology acts in turn to reduce the cumulative value of such data collections. In principle, some of the most exciting questions that such research can help to answer have to do with long-term, secular trends in time use. Yet while studies have been done in many national contexts for several decades, the comparability across most of these investigations is limited or, worse still, unknown. Activity codes are incompatible from study to study; precise interviewing and coding instructions that affect the data are lost; samples are haphazard and unrepresentative. The best that can be said is that there has been rapid methodological improvement, although good time-budget work remains complicated and expensive.
Assuming that they are reliably executed, time budgets by themselves provide extensive but not intensive information: the data are broad but shallow. Their breadth is reflected in the astonishing variety of researchers who can find in them titillating bits of information relevant to their specialties, or useful, if simple, parameter estimates. At the same time, most specialists are dismayed that the relevant information goes no further than it does. Hence, much of recent time-budget work has been moving toward a more multidimensional treatment of the activity itself; it may even involve use of the time budget as a backdrop, or frame of reference, for more intensive inquiry into some theoretically interesting subset of the day’s activities.
Before going further, however, it should be noted that the simple time budget can itself be manipulated in a striking number of ways, each of which has value in one or another specific context. Many studies begin and end with examination of the summary duration of time allotted to the various activities in the code, expressed as an average across respondents. However, it is often useful to modify this information with indications of the prevalence of an activity. Thus, for example, the average duration of working time outside the home is greater in the Soviet Union than in the United States. However, most of the increment is due to a much higher proportion of working women, rather than to long work hours for individuals. Thus both the collective fact and the individual fact are of social significance.
For other inquiries, the timing of activities across the 24 hours of the day is of prime importance. This is true of programmers for the mass media, who need to know when in the day various portions of the populations will be at home and ready to turn on their television or radio sets. Similarly, regional planners and traffic experts collect and use data on the time of day that commuting and other travel take place as well as data on the duration of the journey to work. Studies of shift work by time budgets show the great differences in off-work activity that occur when a person’s work schedule is out of phase with the modal rhythm in the society. Or again, time budgets can be analyzed from the point of view of patterns in the sequence of activities, quite apart from their duration or timing. Finally, people with similar patterns of
activity duration may also differ markedly in the frequency with which the same activities are repeated in a day’s time: some organize their days into relatively large blocks, while others pick up activities and drop them often, either by design or through interruptions from the environment.
Despite the numerous ways in which unadorned time-budget data can be viewed, there is an increasing tendency toward expansion of the information about the activities themselves. It is becoming commonplace, for example, to collect information on where and with whom each activity took place. The locus of activities and their sequence are again of interest in regional planning and have led to conceptualization of the “household activity system” as a part of urban ecology (Chapin & Hightower 1966). Time-budget data designating the “partners” for various activities can be used to test hypotheses concerning variations in interpersonal contacts (Reiss 1959).
Nelson Foote and Rolf Meyersohn have experimented with the collection of further information concerning the affect felt toward each activity reported, the desire for more or less time for it, whether it is seen as work or play, routine or unusual, and whether the activity was initiated by the subject or by others acting toward him. Similarly, they have collected time budgets reflecting more of the nexus of social behavior, by pairing diaries for husbands and wives covering the same day (Foote 1961, p. 171). Somewhat similar analyses attempt to draw in all of the significant social microcosm of the individual—family, neighborhood, and co-workers (Chombart de Lauwe 1956).
While such expansions are helping to enrich time-budget studies, the fact remains that a wide gulf still exists between the manifest activity as it is recorded and the latent functions of the activity for the individual which give it ultimate meaning or significance. The act of repairing an appliance in one’s home may reflect a desire to convert free time into money saved to piece out one’s income, or an inability to secure sufficiently rapid service from outside, or a hobby of tinkering, or some combination of the three. More theoretically satisfying ways of grouping detailed activities into broader classes are often avoided because of just such difficulties. Where ingenious activity groupings are made, there lurks the inevitable suspicion that were the truth known, the same activities might fall in quite different classes for various subjects and that the substantive results may be artifacts of such “forcings” of the data.
It is for reasons of this sort that time budgets seem to have been found less useful for investigations of life style than might be expected on the surface. Activity designations tend to be more incisive about the forms of activities than about their contents. The fact of watching television or having a conversation with a neighbor is recorded, but the types of programs watched or the subjects of informal discourse are rarely catalogued. And variations in what is called “life style” seem thoroughly muted in the process, although some differences do linger on in the forms themselves.
Thus, at one level time-budget data convey an unusual range of implications for practical policy planning and also hold at least some interest for social theory. At the same time they are costly to collect in reliable ways, and without considerable expansion of the texture of information gathered, they often seem disappointingly primitive.
Philip E. Converse
[See alsoindustrial relations, articles onindustrial and business psychology and the sociology of work; Leisure; Planning, social, article onregional and urban planning; time, article on social organization; and the biography of Lund-berg.]
Anderson, Nels 1961 Work and Leisure. New York: Free Press.
Chapin, F. Stuart jr.; and Hightower, Henry C. 1966 Household Activity Systems: A Pilot Investigation. Urban Studies Research Monograph. Chapel Hill: Univ. of North Carolina, Institute for Research in Social Science, Center for Urban and Regional Studies.
Chombart de lauwe, Paul H. 1956 La vie quotidienne des families ouvrieres: Recherches sur les comporte-ments sociaux de consommation. Paris: Centre National de la Recherche Scientifique.
De grazia, Sebastian 1962 Of Time, Work and Leisure. New York: Twentieth Century Fund.
Foote, Nelson N. 1961 Methods for Study of Meaning in Use of Time. Pages 155-176 in Robert W. Klee-meier (editor), Aging and Leisure: A Research Perspective Into the Meaningful Use of Time. New York: Oxford Univ. Press.
Guilbert, Madeline; Lowit, Nicole; and Creusen, Joseph 1965 Problemes de methode pour une enquete de budgets-temps: Les cumuls d’occupations. Revue frangaise de sociologie 6:325-335.
Kleemeier, Robert W. (editor) 1961 Aging and Leisure: A Research Perspective Into the Meaningful Use of Time. New York: Oxford Univ. Press.
Larrabee, Eric; and Meyersohn, Rolf (editors) (1958) 1960 Mass Leisure. Glencoe, 111.: Free Press.
Lundberg, George A. et al. 1934 Leisure: A Suburban Study. New York: Columbia Univ. Press.
Meier, Richard L. 1959 Human Time-allocation: A Basis for Social Accounts. Journal of the American Institute of Planners 25:27-33.
Nakanishi, Naomichi 1963 A Report on the How-Do-People-Spend-Their-Time Survey: An Analysis of Livelihood Time of the Japanese. Tokyo: NHK Radio & TV Culture Research Institute.
Reiss, Albert J. Jr. 1959 Rural-Urban and Status Differences in Interpersonal Contacts. American Journal of Sociology 65:182-195.
Sorokin, Pitirim A.; and Berger, Clarence Q. 1939 Time-budgets of Human Behavior. Harvard Sociological Studies, Vol. 2. Cambridge, Mass.: Harvard Univ. Press.
Szalai, Alexander 1966 Trends in Comparative Time-budget Research. American Behavioral Scientist 9, no. 9:3-8.
United Nations Educational, Scientific and Cultural Organization 1966 The Multinational Comparative Time Budget Research Project. Report for the Sixth World Congress of Sociology, Evian, France. Vienna: European Coordination Center for Research and Documentation in Social Sciences.