Urban Activity Patterns Research Paper

Analyses of urban activity patterns may be traced back to Chapin (1974) and his colleagues, who attempted to explain differences in activity patterns using person and household characteristics, in particular, ‘stage in the family life cycle.’ Another important concept is Hagerstrand’s space–time prism (Hagerstrand 1970). Jones et al. (1983) combined these two approaches and extended them to form an analytical framework, which incorporates both needs and desires that motivate engagement in activities, and constraints that govern activity engagement and travel.

Since engaging in activities implies using time in certain ways, activity patterns are synonymous with patterns of time use. Urban activity patterns have also been examined from time use perspectives. Studies have examined the allocation of time to different types of activities, activity duration, time-of-day characteristics of activity engagement, and sequencing of activities.

Additional elements are spatial aspects of activity patterns, which shall be termed ‘travel patterns.’ Travel patterns have been examined extensively in the field of transportation planning. Of particular interest is trip chaining, which refers to the linking of trips to visit a series of activity locations. (A trip refers to the movement from an activity location to the next activity location. In cases where movement itself is the activity (e.g., walking the dog), the entire movement is a trip.) Also of interest from both time use and travel perspectives are interpersonal linkages, day-to-day variations in activity patterns, and behavioral dynamics.

Characteristics of urban activity patterns are discussed below on each of the aforementioned subject areas. For other reviews of the field and additional references, see Damm (1983), Jones et al. (1983, 1990), Kitamura (1988), Axhausen and Garling (1992), Hanson and Hanson (1993), Garling et al. (1994), Pas (1997), Pas and Harvey (1997), and Bhat and Koppelman (1999).

2. Time Allocation

The total of 24 hours each individual has in a day is allocated to different types of activities. Some activities are mandatory due to prior social commitments (e.g., paid work), some are needed for physiological reasons (e.g., sleep) or for subsistence (e.g., grocery shopping), and some are discretionary (e.g., recreation). There are activities whose location, starting time, and ending time are predetermined, such as paid work. These activities constitute ‘pegs’ around which other activities are arranged.

The amounts of time allocated to sleeping, meals, or personal care are surprisingly similar across data from different regions and times (e.g., see Szalai 1972). Once the population is broken down into segments, data reveal meaningful differences that reflect the characteristics of respective subgroups and regions. For example, American women spend more time on paid work than Dutch women, simply because more American women are employed and more are full-time employees among those who are employed; working American women also spend more time on personal care than their Dutch counterparts.

Time allocation has been studied with theoretical microeconomic models and more descriptive structural equation models (SEMs). A microeconomic theoretic basis can be found in Becker (1965) on which many empirical studies have been based. Because the total amount of time available is fixed, the amounts of time allocated to different types of activities tend to be negatively correlated with each other. For example, as one might expect, those who work more hours spend less time on out-of-home maintenance activities or in-home recreation. A common finding of previous studies indicates that individuals with higher income tend to spend more time on out-of-home recreational activities. Effects of life cycle are evident; individuals living alone and couples without children are more likely to pursue out-of-home recreational activities; single parents and those in extended families are least likely to do so.

3. Activity Duration

The time allocated to a type of activity may be divided into segments. Unfortunately, there seems to be no consensus on how such a segment should be defined. One candidate is the activity ‘episode,’ which shall be defined as a block of time during which none of the dimensions of activities (what is being done, where, with whom, for whom, etc.) change. Another construct is a ‘project,’ which shall be defined as a contiguous block of time in which a sequence of actions is taken to fulfill a purpose. Yet it is typical that the individual engages in multiple activities simultaneously, but not necessarily continuously, that serve a multitude of purposes, making the definition of projects difficult, or arbitrary, to a large extent.

Some projects have fixed durations while others have variable durations that are affected by a number of factors. For example, the project duration at an out-of-home location is positively correlated with the length of the trip to the location. Commute duration affects people’s use of time; people with commutes longer than 60 minutes tend to spend most of their evenings indoors. Reductions in commute time induce more travel, i.e., a portion of travel time savings is invested in more travel. Consistent with the results of time allocation studies, older individuals and individuals with lower incomes tend to have longer homestay durations.

4. Time Of Day Characteristics Of Activity Engagement

Many activities repeated daily have regular starting and ending times, for physiological and institutional reasons. Sleeping, eating meals, and work activities are repeated highly regularly. Californian and Dutch time-use datasets indicate that personal care has a peak in the morning, then distributed throughout the day, with another peak in the late evening. Activities such as childcare, on the other hand, have no peak. Running errands is distributed over a period between 8:00 a.m. and 10:00 p.m. with no clear peak. Social activities are distributed over a similar period, with higher intensities in the evening. Engagement in entertainment, reading, and TV viewing tends to increase toward the end of the day and peaks in the evening. These time-of-day characteristics of activity engagement have been examined with the concept of time-varying utilities of daily activities, and time-of-day effects have been quantified. Past studies have also found that characteristics of activity location choices differ by time of day.

5. Activity Sequencing

A set of multiple activities tends to be sequenced in certain ways by individuals. Those activities that are more mandatory appear to be pursued first as people tend to pursue out-of-home activities in a day in the following order: serve-passenger activities, personal business, shopping, then social recreational. Similar tendencies are found within a trip chain. These sequencing patterns are correlated with the time-of-day characteristics of activity engagement discussed above.

Activities are engaged under a set of constraints. Most important constraints are work schedules, which typically define three space–time prisms for a worker. The intensity of constraints varies across individuals. For example, the concentration in time of commute trips is interpreted in a study as an indicator of how constrained trip making is, and it is noted that unmarried women with children have the least flexible, and singles have the most flexible, work trips. Several studies have attempted to explicitly incorporate constraints into models of activity engagement.

6. Activity And Travel

A trip is generated each time the activity location changes. Although the spatial and temporal characteristics of trips have been extensively studied, the mechanism of trip generation is not well understood. A theoretical model of shopping trip frequency has been developed by applying the inventory theory. This theoretical framework, however, is not applicable to all types of trips with nonwork purposes. The concept of time-dependent utility has been adopted to show when a trip will take place, and later some empirical evidence was offered in support of the model. These are among a very few examples in which the mechanism of trip generation is directly addressed.

From microeconomic viewpoints, more travel is done—by making more trips, traveling longer distances, or both—when the cost of travel is less. Then individuals residing in areas with higher accessibility to opportunities should be traveling more. Empirical evidence is not conclusive on this point, however. For example, a study found the ‘degree of urbanization’ is not associated with travel; another study reports that household location classification variables based on density measures are not significantly associated with the number of stops by nonworkers. On the other hand, very significant effects of GIS-based accessibility measures are reported in a study.

Attempts have been made to treat daily travel behavior in its entirety. Approaches taken in these studies include: applications of discrete choice models and applications of mathematical programming concepts. The former treats daily travel behavior as the choice of a travel pattern from among all possible alternative patterns. With the spatial and temporal dimensions, the number of alternative patterns can be astronomical, whose enumeration may impose computational difficulties. In addition, it is not compelling that the tasks of enumerating and evaluating numerous alternatives are manageable with human cognitive capacity. Likewise, treating the decision process underlying daily behavior as a mathematical programming problem does not appear to offer realism either. Furthermore, if viewed as an optimization problem, the decision process deals with an extremely complex problem of formulating and finding a solution for, leading to prohibitive computational requirements. An alternative approach is to decompose the decision associated with a daily travel pattern into a series of sequential decisions. Sequential approaches offer a better representation of cognitive processes underlying scheduling behavior. Yet they may not adequately depict individuals’ responses to changes in the travel environment.

7. Typologies Of Activity–Travel Patterns

Yet another approach to daily travel patterns is to develop typologies of travel patterns. An early work proposed a seven-group classification of travel patterns, obtained by applying a grouping procedure to factor scores that comprise measures of travel patterns, e.g., number of trips and total travel time. A later study developed ‘geometrical similarity indices’ based on a hierarchy of activity pattern attributes, and proposed travel pattern typologies that include five patterns of trip chaining. Application of pattern recognition theory and a variety of similarity measures can be found in the literature.

8. Urban Travel Pattern Characteristics

There is a substantial body of literature on urban travel patterns. A majority of individuals make ‘simple’ daily patterns of visiting one out-of-home location with two trips, i.e., a trip from home to the location, then a trip from the location back home. According to the 1990 National Personal Transportation Survey (NPTS) data, 30.4 percent of 48,385 respondents were reported to have made just two trips on the survey day, 8.6 percent three trips, 15.3 percent four trips, and 23.1 percent more than four trips. On the other hand, 21.2 percent of these respondents reported no trips at all. The average number of trips per person per day is 3.09; the average for those who made at least one trip is 3.92.

Gender roles and family life cycle are strongly associated with travel patterns. A number of studies have found that women engage in shopping more than men. An analysis of a 1990 dataset from the San Francisco Bay area indicates that women make about 75 percent more grocery trips than men, regardless of race/ethnicity, do more child chauffeuring, and make more household-serving trips.

The organization of nonwork travel, especially linking work and nonwork trips, varies with household structure. For example, singles tend to be away from home; two-person households have a positive propensity to conduct activities out of the home; and households with children tend to have more nonwork activity time, more nonwork trip chains, and more travel time to nonwork activities.

Household auto ownership is another factor that is strongly associated with household members’ travel patterns. It is associated not only with the choice of the means of travel (or ‘travel mode’) but also with the number of trips itself. Study after study has shown a negative association between residential density and auto ownership, while household auto ownership is the principal explanatory variable of household trip generation that has traditionally been used in transportation planning studies.

9. Trip Chaining

Trip chaining is another subject area that has been extensively studied (e.g., see Hanson 1979, Golob and Golob 1981). Early models of trip chaining are Markovian that attempt to replicate linkages between activities pursued at successive out-of-home locations. This was later extended to include a continuous time dimension, and the stationarity and history dependence of activity transitions were statistically examined.

Many studies have indicated that individuals with higher incomes tend to chain trips more frequently and make complex trip chains. There are, however, results that income is not associated with forming complex commute chains, or indications that households with higher incomes have smaller propensities to chain trips.

Females appear to be more active than males, with more trips, more trip chains, and more trips per trip chain. On household structure and trip chaining, it has been observed that households of young working adults without children tend to have chains involving work trips; households with preschool children tend to make simple one-stop chains and simple commute trips; households with school-age children are inclined toward complex chains; while elderly households tend to have simple chaining patterns. Life cycle is also associated with the generation of serve-passenger trips, and making more serve-passenger trips implies more trip chains.

A study of the 1983 NTPS data indicates that nonwork travel accounted for just over half of all person trips in the a.m. peak and two-thirds in the p.m. peak, and nonwork trips grew considerably faster than work trips between 1977 and 1983. This may be due to the linking of nonwork travel to commute trips. Importance of commute trips as a base to form trip chains has been noted by several researchers. Reductions in household size coupled with increases of multiworker households result in a greater tendency of linking nonwork trips to work trips.

10. Interpersonal Linkages

Most analyses of urban activity and travel patterns are concerned with the behavior of each individual. Although the focus on life cycle stage is in part based on the understanding that interactions among household members affect each member’s activity and travel, characteristics of interpersonal interactions have been relatively unexplored. Subjects for investigation include how resources are allocated to household members, how tasks are assigned or performed on somebody else’s behalf, and how activities are pursued jointly. Also of interest is how interpersonal interactions motivate the engagement in an activity, either jointly or separately. Such interactions are expected to take place through various social networks.

Inter-relationships among household members may be classified into substitutable relationships, complementary relationships, and companionship relationships. A study indicates that relations between household members are mostly complementary. Evening space–time paths of male–female couples have been examined and yielded findings that those with preschool or school-age children have higher propensities toward evening out-of-home activities, and young couples without children are oriented toward joint activities, often after meeting each other outside the home. Studies have shown that male travel demand is sensitive to female work and maintenance activities.

11. Day-To-Day Variability

Another limitation of most of the studies is that they are concerned with activity and travel patterns for just one day, mostly a weekday. This reflects the fact that most datasets are based on survey periods of just one day. Such data contain patterns on a randomly sampled day, carrying no information on how typical or unusual that day might have been.

Examining day-to-day variability is important for several reasons. First, without information on day-today variability, it is impossible to determine exactly how much of the interperson variability in travel patterns is genuine, and how much is the artifact of intraperson, day-to-day variability. Likewise, it is not possible to detect how much behavior has changed over time. Second, statistical analysis of one-day data may not yield a correct depiction of behavioral change if it is history dependent. Third, one-day data do not indicate how many individuals exhibit certain behaviors over a span of time. For example, suppose oneday data indicate that 15 percent of commuters travel by public transit. This may mean a specific group of commuters, which comprises 15 percent of all commuters, travel by public transit every day, or every commuter commutes by public transit 15 percent of the time.

The level of repetition and variability in daily travel patterns has been studied using 35-day travel diary records, yielding findings that weekday and weekend patterns are different, and each individual has more than one typical weekday pattern. It is further shown that the level of repetition varies from population subgroup to subgroup. It has been reported that role related constraints and commitments are associated with the variability in the individual’s travel patterns. A study postulates a two-stage process: weekly behavior is selected first, then daily behavior is selected. Employed people with low incomes tend to select ‘simple’ weekly travel activity patterns. In another study, a two-stage model system is adopted in which a ‘latent’ daily pattern is selected, then travel time expenditure is determined.

12. Behavioral Dynamics

Most studies of urban activity patterns assume that patterns are stable or in equilibrium. Underlying this is the assumption that individuals adjust their activity patterns immediately and precisely whenever a change takes place in their travel environments or in their preferences (Goodwin et al. 1990). This of course is a very strong assumption. In fact, responses to changes often involve time ‘lags.’ Learning, or trial-and-error search for adaptation schemes, takes time, also producing lags. On the other hand, behavioral responses may precede changes in the environment because of individuals’ planning actions, resulting in ‘leads’ in responses.

It is often the case that there are no responses to changes because of habit and behavioral inertia. Some of the issues addressed in day-to-day variability pertain to behavioral dynamics. The accumulation of empirical results is still in progress in this subject area, and it is premature to assess common dynamic characteristics of activity and travel behavior at this time. It is noted that there are pieces of empirical evidence which demonstrate the presence of behavioral intertia, show discrepancies between longitudinal elasticities and cross-sectionally evaluated elasticities, and demonstrate dynamic properties in many aspects of activity and travel behavior.

Bibliography:

1. Axhausen K W, Garling T 1992 Activity-based approaches to travel analysis: Conceptual frameworks, models, and research problems. Transport Reviews 12(4): 323–41
2. Becker G S 1965 A theory of the allocation of time. Economic Journal 75: 324–41
3. Bhat C R, Koppelman F S 1999 A retrospective and prospective survey of time-use research. Transportation 26(2): 119–39
4. Chapin F S 1974 Human Activity Patterns in the City: Things People Do in Time and Space. Wiley, New York
5. Damm D 1983 Theory and empirical results: A comparison of recent activity-based research. In: Carpenter S, Jones P M (eds.) Recent Advances in Travel Demand. Gower Publishing, Aldershot, UK
6. Garling T, Kwan M-P, Golledge R G 1994 Computationalprocess modelling of household activity scheduling. Transportation Research B 28B(5): 355–64
7. Golob J, Golob T F 1981 Classification of approaches to travelbehaviour analysis. In: Travel Analysis Methods for the 1980’s, Special Report 201, Transportation Research Board, Washington, DC
8. Goodwin P, Kitamura R, Meurs H 1990 Some principles of dynamic analysis of travel behaviour. In: Jones P (ed.) Developments in Dynamic and Activity-Based Approaches to Travel Analysis. Avebury, Aldershot, UK, pp. 56–72
9. Hagerstrand T 1970 What about people in regional science? Papers of the Regional Science Association 24: 7–21
10. Hanson S 1979 Urban travel linkages: A review. In: Hensher D A, Stopher P (eds.) Behavioural Travel Modelling. Croom Helm, London
11. Hanson S, Hanson P 1993 The geography of everyday life. In: Golledge R, Garling T (eds.) Behavior and Environment: Psychological and Geographical Approaches. Elsevier Science, Amsterdam, pp. 249–69
12. Jones P, Koppelman F, Orfueil J P 1990 Activity analysis: State-of-the-art and future directions. In: Jones P (ed.) Developments in Dynamic and Activity-Based Approaches to Travel Analysis. Gower Publishing, Aldershot, UK
13. Jones P M, Dix M C, Clarke M I, Heggie I G 1983 Understanding Travel Behaviour. Gower Publishing, Aldershot, UK
14. Kitamura R 1988 An evaluation of activity-based travel analysis. Transportation 15: 9–34
15. Pas E I 1997 Recent advances in activity-based travel demand modeling. In: Texas Transportation Institute (ed.) Activity-Based Travel Forecasting Conference, June 2–5, 1996, Summary, Recommendations and Compendium of Papers, Travel Model Improvement Program, US Department of Transportation, Washington, DC, pp. 79–102
16. Pas E I, Harvey A S 1997 Time use research and travel demand analysis and modelling. In: Stopher P, Lee-Gosselin M (eds.) Understanding Travel Behaviour in an Era of Change. Pergamon, Oxford, UK, pp. 315–38
17. Szalai S (ed.) 1973 The Use of Time. Mouton, The Hague