Environment and Planning, 1971, volume 3, pages 353-368

Space-time budgets and activity studies in urban

geography and planning t

J Anderson
Department of Planning and Urban Design, Architectural Association School of Architecture, London
Received 17 May 1971

Abstract. This paper attempts to provide a general perspective on the potentialities and limitations
of research into individual activity patterns. The most common types of activity study and data-
collecting device are briefly described, and problems of data collection and analysis are outlined.
The main focus is on the value of 'space-time budgets' in household surveys, assessed with respect
to two contrasting aims: (1) deriving 'behavioural' postulates on which geographic theories of
spatial structure might be based, and (2) planning spatial structures to suit the behaviour patterns
and aspirations of different types of individuals and households. Activities occur in a 'space-time'
continuum', and there are temporal regularities inherent in spatial patterns. It is argued that
studying activity patterns in terms of 'space-time locations' can throw light on pressing socio-
spatial problems.

1 Introduction
A time-budget is a systematic record of a person's use of time over a given period.
It describes the sequence, timing, and duration of the person's activities, typically for
a short period ranging from a single day to a week. As a logical extension of this
type of record, a space-time budget includes the spatial coordinates of activity
locations.
Studies using these devices do not constitute a unified research field. The purposes
of such studies vary widely. In the English-speaking world, time-budgets have been
used mainly in market-research for the mass media and in non-spatial sociological
studies of life-styles and leisure (Converse, 1968). Mead's (1958) paper on seasonal
variations in rural life in Finland was one of very few instances of time-budgets being
used in geographic research prior to 1960. More recently, however, work by Meier
(1962), Chapin (1965), and Hagerstrand (1970a) has given greater impetus to
spatially-oriented research on daily activity patterns. Space-time as well as time-
budgets are now being used, and mainly in urban rather than in rural studies. The
use of a related device, the travel diary, by Marble (1959, 1967, 1970) and by
Nystuen (1959, 1967) also contributed to the growing interest in relationships
between uses of space and time.
There is considerable variation in the underlying methodology and purposes
associated with the use of activity records, even within the field of urban spatial
research. Meier, for instance, sees their potential, in practical rather than theoretical
terms, as a means of gauging the general quality of urban life and overall efficiency of
urban systems. Chapin, taking a more microscopic approach, sees activities as the
outcome of choices which reflect people's values, while Hagerstrand, in contrast,
emphasizes the constraints imposed by urban environments. He is mainly concerned
with how people's activities are 'canalized' through urban space-time, for example,
with how city size affects daily patterns of life. As significant as their differences,
however, is the fact that for all three, urban planning is a central consideration.

t This paper was presented in a shortened version at the Institute of British Geographers' meeting
on 'Space-Time Budgets in Urban Analysis', at Cambridge University, 7 November 1970, and
published as Discussion Paper 40 in the London.School of Economics Geography Department series.
354 J Anderson

Policy considerations are also central to the transportation economists' work on the
value of time, and trade-offs between alternative travel modes (for example, Foster
and Beesley, 1963; Quarmby, 1967); in order that transport investment and pricing
can be related to travel-/7rae savings, time is evaluated in monetary terms.
This cash-conversion procedure is perhaps peripheral to more comprehensive studies
of the spatial organization of diverse activities and activity-sequences. Time is itself a
quantitative measure highly appropriate in studying social organization. As a
pervasively awkward scarcity, it is perhaps the main rival to money in our society
(Moore, 1963); in many situations 'time is money'. It is not simply a quantity to be
'spent', 'saved', or 'wasted'; we also have to 'keep time', meet deadlines, and order
our activities in particular sequences, simultaneously taking account of the amounts of
time required to get from place to place. Time is a 'one-directional dimension' to
which we refer in organizing our activities, and it can serve as a scale on which most,
if not all, activities and their interrelationships can, in principle, be measured. It can
be used to create typologies of activities, people, institutions, and environments,
which are directly appropriate to studies of accessibility and spatial behaviour
(Carlstein et al., 1968). For comparative purposes, time (unlike money) has the
advantage that everyone who remains alive during a given period has exactly the same
amount to 'allocate'. This is not to disparage monetary units (either for spending or
in studying behaviour)—lack of cash is a major constraint on time allocation—rather
it is to suggest that temporal units have considerable potential in spatial studies,
without being converted to cash terms. Indeed, economists have recently begun to
consider all time (non-work as well as work hours) as a 'scarce resource' (Burenstam
Linder, 1970); and attempts by them (for example, Burenstam Linder, 1970;
Becker, 1965) to develop a general theory of time allocation may eventually prove
useful in constructing deductive spatial theories. The concept of trade-offs between
time uses is, at least potentially, very relevant to spatial study. However (as is argued
in Sections 5 to 7 below) while being alive to such theoretical implications, it is
perhaps more realistic at present to consider time—and the space-time budget device
—as primarily of use in descriptive or 'inductive' studies of spatial behaviour.
When a device or technique is the focus of attention there may be difficulties in
going beyond purely technical considerations. Formidable problems of data collection
and analysis are the most immediately obvious common element in time-budget
studies. However, these problems will be discussed only in general terms (Section 4),
and the main types of activity study and the most usual devices for collecting data
will merely be outlined (Section 3). The emphasis is on the value of activity studies
in human geography, the potential of space-time budgets in such studies, and the
relevance of this type of study to urban planning (Sections 2, 4 to 8).

2 Time and space

Curry (1967a) describes the recent history of geography as "the adoption of methods
developed elsewhere and the gradual appreciation of the difficulties of spatial
analysis", Time-budgets are such an adoption, and while their use, involving as it
does explicit treatment of the temporal dimension, may compound analytical
difficulties, it may add to our understanding of important spatial problems.
Time-budget research developed out of 19th century studies of urban life and
poverty (for example, Engels, 1892), and was later influenced by 'time-and-motion'
studies. The first large empirical survey of 24 hour time-budgets was carried out in
1924 on Moscow workers, as part of national economic planning, and from it
developed a substantial tradition in Eastern European social research. In 1963, time-
budgets formed part of an official census in Hungary. Their use in the West has been
more sporadic; the largest surveys have been geared to radio and television
Space-time budgets and activity studies 355

programming, but the focus has most frequently been on general leisure and life-
styles (for example, Lundberg et al., 1934; Sorokin and Berger, 1939), an emphasis
continued in the present UNESCO sponsored cross-cultural time-budget study
involving over ten countries (Szalai, 1966; Converse, 1968). Much of this work
gives, at best, only partial treatment to spatial locations (see Anderson, 1970).
The French sociologist, Chombart de Lauwe (1960), has used time-budgets to show
the effects of locations and differential journeys to work on family life and welfare,
and in architecture, activity studies have provided a basis for designing interior
layouts. In anthropology, time allocations have served as indicators of cultural
change, and production time has been taken as a measure of cost-value in non-
monetary economies (Firth, 1967). In 'mainstream' economics it is only recently
that all time (that is, 'free' or non-work as well as work-time) has been considered
explicitly as a scarce resource; according to Burenstam Linder (1970), economists
have traditionally made the implicit assumption that consumption, in contrast to
production, was atemporal, whereas, in fact, like production, it has to be
accommodated in a time-table. In modern production-oriented 'consumer society'
temporal scarcity seems to be increasing (see Anderson, 1971).
Modern citizens organise their daily lives by reference to clock and calendar, and
in accordance with deadlines over which they often have little control. Conceptions
of time vary for different activities and social groups (Gurvich, 1964); and Quarmby
(1967) in discussing the monetary value of travel time suggests that waiting time may
have to be costed at three times the rate for time spent actually moving. In peasant
societies, personal time indicators (for example, the 'biological clocks' of hunger and
tiredness) may have more relevance than impersonal clock-time; in Greek peasant
tradition "to hurry is to forfeit freedom", the peasant is "intent on passing the time,
not on budgeting it" (Mead, 1954). We too are guided by 'biological clocks', but
these are frequently overruled by mechanical clock-time (and not without danger to
health and welfare—see Wilkins, 1968). If the variable natures of time are taken into
account, it loses some of its convenience for activity-study purposes; considering its
paramount importance in industrial societies, clock-time may provide an adequately
'objective' standard. Mumford (1934) chose the clock rather than the steam-engine
as the 'key machine' of the industrial age. The increased specialisations, spatial
separations (for example, of home and work) and interdependencies, which are a
general feature of industrial growth, require more precise time scheduling and tend to
reduce flexibility in time allocation (Anderson, 1961). Space, distance, and
accessibilities acquire new meanings as the importance attached to time increases.
Human behaviour, although highly variable, does display some marked temporal
regularities, because of physiological, physical and social constraints. Activities occur
in a space-time continuum: there are temporal regularities inherent in spatial
regularities, and temporal r h y t h m s obviously vary over space. The c o m m o n practice
in geography of collapsing the time dimension may simplify spatial analysis, but in so
doing we may ignore what Nystuen (1963) has called 'space-time tensions': "When
time is short, space is conserved. When an activity has a deadline associated with it,
congestion in space is likely." Urban rush-hour congestion is ample proof of the
latter point; and distance minimization can often be expressed realistically as time
minimization. For the individual, residential location, spatial mobility, accessibility
levels, and time-use patterns form a complicated tangle of interrelationships. If
geographers are to untangle some of these, without unduly distorting them,
questionnaire surveys combined with activity records seem essential (see Section 7
below).
356 J Anderson

3 Types of activity and activity record

Detailed activity records describe the whole time-use pattern, or preselected types of
activity (for example, travel), of individuals over a particular period. Respondents
record their activities in a diary, or a temporally-structured questionnaire is used
during interview (for example, the 'day-after' method). Whether in questionnaire or
diary form, activity records can be classed in four categories:
1 time-budgets and
2 space-time budgets (as already defined) may have additional data on activity
participants, travel modes, and the degree of recurrence, compulsion, or constraint
associated with the activities. Both, in principle, cover all time allocations
throughout the specified period, and the shorter term, 'time-budget', is sometimes
used for both.
3 contact records cover face-to-face and other preselected types of interperson
communication, while
4 travel records typically describe the origins, destinations, purposes, modes, and
times of trips. In some (for example, the London Traffic Survey, 1964, 1966),
only vehicle trips are included and the occurrence of walking trips can produce
discontinuities in the record of the respondent's space-time path. Activities at
'fixed' locations or stations are not described, although some of them may be
inferred from trip 'destination' and 'purpose'.
Chapin and Hightower (1965) suggest that time-budgets are required for land use
planning, as travel records are in transportation planning; in fact, the 'origin and
destination' data of the latter are also contained in a space-time budget, which is the
most general and comprehensive of the four types of activity record.
Urban activity systems may be divided into three general classes depending on their
principal nodes: firms, institutions, and residences (Chapin, 1965). Contact diaries of
personnel (especially decision-makers) are, for instance, being used in Sweden to discover
the communications and spatial connections of firms (see Anderson and Goddard,
1969); and space-time budgets are being used to discover the links between a large
institution, Bedford College, London, and the surrounding urban area (Nichols, 1969).
Residence based surveys are more diverse in their objectives, and perhaps more
'basic' in that all sections of the population—young, old, working, not working—can
in principle be covered. They are sometimes more difficult: sampling problems may
be greater, activities and the factors influencing them are usually more varied-
households are less 'institutionalized' than firms. However, a household is an
institution of a kind. It can be considered an 'activity sub-system', in which the
members of the household influence and constrain to varying degrees each other's
activities and movements. Some activity surveys cover all 'adult' members (such as
all those over 11 years); others, because of the difficulties involved, make do with
just one adult, usually the household head or spouse.

4 Data collection and analysis

But for some formidable problems in the collection and analysis of space-time data, we
might agree with Minkowski, who in 1908 wrote: "Henceforth space by itself, and
time by itself, are doomed to fade away into mere shadows, and only a kind of
union of the two will preserve an independent reality."
The data are expensive to collect. Unless 'day-after' methods are used, completed
diaries have to be retrieved from respondents. If the 'day-after' approach is used it
is difficult to interview all adult members of a household, or, because of fading
memories, cover more than one preceding 24 hour period. It is not easy to check for
accuracy, or to check that recorded activities are in general typical of recurring
behaviour patterns. It is necessary to get information (by questionnaires) on
Space-time budgets and activity studies 357

activities which occur irregularly or follow a longer cycle than the activity record
period, although seasonal variability presents difficulties. The willingness of
respondents to record activities falls off fairly rapidly, whether or not payment is
involved—about three days seems a critical level for relatively comprehensive diary
records. However, a short activity record period facilitates comprehensive treatment
of important daily and weekly rhythms.
Research design and sampling are particularly important because of these problems.
The activity classifications in sociological studies are typically 3-digit and very
detailed. For spatial studies a simpler and more flexible classification system may be
more suitable—for example, the first digit might specify the 'station' (home,
workplace, shop, other person's home, and so on) while the second (and third)
describes the nature of the activity (working, purchasing, travelling, and so on);
behaviour can be analysed in terms of each component separately and in terms of the
whole classification system. Recorded addresses must be translated (laboriously) into
suitable locational terms, such as a grid system reference, perhaps with a variable grid
size depending on the use made of different parts of the overall activity space.
Standardized 'time blocks' are also problematical, especially at the recording stage
where they are likely to impose extraneous 'structure' on the time pattern of
activities.
Space-time activity patterns may be influenced by a wide range of factors, some of
which (for example, age, occupation, household composition, or residential location)
must a priori initially be taken as particularly important and used in sample
stratification. To ascertain the influence of one variable, such as home location, a
whole range of other factors should ideally by held constant. Thus households who
intend moving to new residences might be surveyed before and after they move
(Blight, 1970) although observed activity changes might be due partly to additional
factors, for example, premeditated choice, decreased knowledge of the 'environment',
or both (see Section 8 below). This implies an important point: to be really useful,
activity analysis should be linked to other types of analysis (such as attitudes to
environment and accessibilities) and this demands careful integration of diary and
questionnaire designs.
Problems of analysis are probably solved less easily than those involved in data
collection (Hurst, 1969). They can perhaps be clarified by distinguishing between a
substance language and a space-time language. The latter—a locational coordinate
system of four dimensions (or three, if the vertical axis of space is ignored)—is
associated with the so-called 'physicalists', who advocated the unification of science
through use of the language of physics (Neurath, 1944). The non-spatial substance
language describes characteristics or properties (occupations, types of activities) of
respondents or other individuals, while their locations or positions are described in
terms of a space-time language. Writing in a non-geographical context, Wilson (1955)
points out that individuation requires the use of space-time coordinates; if
individuals are defined only in substance terms they "lose part of their essence".
This part is particularly essential to geographic study. Harvey (1969, p.217) writes
that the ordering of geographic data "amounts to the difficult logical problem of
working with two different language systems in the same context"—a problem
increased by the fact that geographers usually deal with both discrete and continuous
data. He cites regionalization as a case where different languages have been mixed up
to produce confusion and controversy.
The problem of using both languages together lies at the heart of space-time
budget analysis. The substance language is itself multi-dimensional (for example,
different types of respondents, activities, non-locational constraints, attitudes), and
adding the temporal dimension to the two spatial dimensions introduces further
358 J Anderson

complications. Hagerstrands' approach, where activity schedules are run through a

computer model of a space-time environment in order to evaluate the constraints
imposed by that environment (see, Anderson and Goddard, 1969), is perhaps the
most comprehensive response to these dilemmas, and is in the physicalist tradition.
It focuses on the 'outer limits' within which behaviour can take place emphasizing
'negative determinants' of behaviour rather than 'positive' factors such as the
attitudes, motives, and choices which contribute to shaping activity patterns
(Section 7 below). The latter emphasis is more common in social science, and the
more common response to space-time data sets has been to abstract from them and
perform one or more separate analyses.
These simplication procedures can be divided into three categories:
a. The simplest procedure is to collapse the coordinate locational language to a
substance form. Organizational aspects of activities—timing, sequence, and spatial
location—are lost, and time spending becomes (like money spending) a
characteristic of respondents, and, in aggregate, of population groups and areas
[for example, Marble's (1959) analysis of travel durations and frequencies and
Meier's (1962, p.53) 'index of urbanity' based on standardized durations of
specified activities in different types of settlement].
b. The timing and/or sequence of activities are made explicit, but their spatial
organizations are not. Thus traffic volumes are plotted over a 24 hour period
(London Traffic Survey, 1964); and sequences of linked activities, or 'activity
bundles', may be isolated, though activity transition probabilities vary depending on
the time of day (Hemmens, 1966; Nystuen, 1967; Brail, 1969).
c. The spatial organization of activity patterns is explicated, but details of temporal
organization are not. Particular spatial linkages and activity locations within the
overall activity space may be mapped for the whole of the activity record period,
or for particular times during it (Chapin and Hightower, 1966; Nichols, 1969). In
addition, a range of more general types of spatial analysis, developed mainly in
other contexts, might be applied to space-time budget data. For example, see
Morrill and Pitts (1967) on fitting distance-decay functions to spatial contact data;
also Moore (1970); and for a graph theory approach to 'functional distance'
(that is, a measure of the 'distance' separating any two nodes which reflects the
net effect of their site and location properties on their interaction propensity),
see Brown and Horton (1970).
All three general categories of analysis, either separately or in various combinations,
may be relevant to geographic study. However, obtaining comprehensive summary
measures of the overall pattern (including sequencing) in space-time budgets remains
a major problem. They might be derived using factor analysis or other means
(Cullen, 1970), and could, ideally, be correlated with other variables covering, for
example, characteristics of respondents and their residential situations (see also,
Burton and Noad, 1968, Chapter 3).

5 Activity records and behavioural geography

In a very literal sense, residential surveys using space-time budgets are a part of
behavioural geography. This, crudely stated, focuses on two related aspects: peoples'
overt spatial behaviour, and their perceptions of their physical and social
environments (Gould, 1969). Space-time budgets record overt spatial behaviour
(moving and stationary), and they can be indirectly related to environmental
perceptions (via questionnaire data) as these interact with overt activities.
Referring to contact and time budget diary research by Swedish geographers,
Haggett (1968) notes that the derivation of ideas from industrial sociology and
organization theory represents a shift away from more traditional types of physical
Space-time budgets and activity studies 359

and biological analogy. Daily life patterns as well as firms and institutions may be
studied in terms of linkages, systems, and organization (Hagerstrand, 1968). Ecology
and, more recently, ethology are other sources of inspiration. Baker (1969), Eyles
(1970), and others have related concepts of animal behaviour, such as territoriality,
to human geography. However, within geography there has been relatively little
testing of such analogies, and hence there is considerable scope for divergent opinions.
[For example, Doherty's (1969) countering of Webber's (1964) suggestion that
territoriality is important for the working class but not for the more mobile
'intellectual elites'; the more mobile may have strong attachments to particular
spaces, that is, to a spatially discontinuous territory.] As far as human behaviour is
concerned, ethological analogies have been most fruitful in small group ecology on a
very limited spatial scale, for example, Sommer's (1969) study of behaviour inside
rooms and its implications for better interior design. Analogous leaps from this scale, or
direct from animal behaviour, to the spatial scale of the neighbourhood or city can be
dangerously misleading. Calhoun's famous rats and mice do not actually tell us
anything about high-density urban living, though their behaviour might suggest
fruitful lines for empirical research into human behaviour (see Calhoun, 1963).
Spatial scale is an important factor in the relative success of ethologically inspired
ecology studies of small-groups. Human behaviour can be directly observed for
continuous periods, and room environments can be experimentally controlled.
[Anthropologists have also used direct observation, but again only for small g r o u p s -
see, Lewis (1959).] For behavioural processes at the neighbourhood or city scales, in
which geographers are more interested, direct and continuous observation is usually
impractical, if not impossible. It has been used by Carlstein et al (1968) to construct
peoples' activity patterns for part of a day, but the study area was a small village and
a well-organized team of observers was required. It was used in a small research
project in Notting Hill, London, but the focus was on the use made of streets at
particular times of the day and week, and information on those observed was lacking
(Arias and Diaz, 1970).
The main quality of space-time diaries is perhaps that they record behaviour
patterns which are not directly observable because of their spatial and temporal
extent. Equally important is the fact that the identity of individuals over time is
preserved—they are not 'commuters' or 'shoppers', or 'leisure-seekers' in isolation
from their other roles. While particular activities may be singled out for analysis, they
can be treated in the context of the respondents' time and space uses throughout the
record period. In a questionnaire study of over 500 households, Clarke (1968) found
that less than half his sample went to the nearest available shops to their home. He
concluded that regression models were not very helpful in explaining consumer
behaviour because of multi-purpose trips and because people visit shops from bases
other than their place of residence. Space-time budgets provide detailed information
on this type of activity organization; they provide an overall behaviour context
within which particular activities can be viewed realistically.
Nystuen (1963) identifies distance, direction, and connection as the three
'fundamental spatial concepts'. Using space-time budgets, distance can readily be
expressed in terms of time, which is often the way people perceive urban distances,
and connection can be given more comprehensive treatment than is perhaps usual in
geography. Haggett (1968) has labelled this emphasis in the study of cities as contact
structures', the emphasis has shifted from distance to contacts, from Euclidean
geometry to graph theory and topology; there has been increased interest in
discontinuous and anisotropic properties of urban space (as measured in terms of
time, costs, or perceptions), and a corresponding dissatisfaction with the assumptions
of classical location theory. It seems highly debatable, however, whether behavioural
360 J Anderson

geography, and in particular the use of space-time budgets, can directly contribute to
improving location theory—as opposed to underlining its already obvious
shortcomings.

6 Spatial behaviour, spatial structure and theory

In revealing the inadequacies of location theory, behavioural studies also reveal the
difficulty of improving on its postulates, or of matching its formal elegance. The
'harmonies' of static equilibrium are indeed 'soothing' (Robinson, 1964, p.70), and
that predictable robot 'Economic Man' is, like Santa Claus, a comforting figure. As
Adams (1968) says, most attempts to make him more realistically human "have
become hopelessly bogged down in the complexities of reality". King (1969) finds
little evidence in behavioural studies that "geographic process theory can be
formulated ... to yield statements of spatial structure as logical 'outputs'".
This is not altogether surprising, for spatial behaviour and spatial structure form a
'circle of causality'. Observed behaviour is in part determined by the structure of the
environment in which it occurs; and Rushton (1969), following Curry (1967b),
argues that parameters of behaviour in a particular environment (such as, empirically
derived distance-decay functions) are therefore "not admissable as a behavioural
postulate in any theory". Such a postulate should incorporate the 'rules' of spatial
choice which underlie behaviour patterns, irrespective of the particular spatial
environment within which the behaviour occurs. Christaller's postulate that
consumers patronize.the nearest place offering a required item is a logically admissable
'rule', but consumers frequently 'disobey' it (Nystuen, 1959; Clarke, 1968). If more
realistic rules of spatial choice are to be found through behaviour studies, geographers
have to break into the 'circle of causality'. Rushton (1969) proposes the concept of
'revealed space preferences' as a means towards this end, and he notes that in
economic consumption theory the spatial distribution of shops is not considered a
significant variable. However, it is significant when the choice is between alternative
shopping locations rather than between different commodities, and Rushton concludes
that his spatial 'preference structures', while they may have more generality than
spatial systems, are not independent of the spatial systems in which they are derived.
This approach might eventually provide useful postulates, but so far it provides only
further evidence of the elusiveness of general 'rules' which are 'independant' and
realistic.
There is little reason to suppose that the use of space-time budgets would be any
more successful in finding these rules. The notion of trade-offs between time
allocations, and between time and space preferences, is conceptually appealing, and
the analysis of behaviour patterns in a wide variety of environments might produce
interesting regularities. To paraphrase Wolpert (1965), the understanding of spatial
behaviour requires determining the constants in space-time patterns and the
distinguishing of these from the variables. But the variables might well prove at least as
interesting as the constants, and, in the absence of spatially-oriented research into
comparative time-use, it is premature to suggest that the constants could be useful as
a basis for a deductive location theory.
Discussing behavioural models, Cox and Go Hedge (1969) distinguish between
(a) the search for behavioural postulates and (b) the relating of actual behaviour to
the spatial environment in which it takes place. The latter focus would seem the
more immediately rewarding as far as space-time budgets are concerned. Space-time
research should not be divorced from theoretical considerations, but there are other
criteria of its usefulness apart from its contribution, or non-contribution, to location
theory. In the short time perspective typical of time-budget studies, the influence of
spatial structure on spatial behaviour is more important than the converse. Urban
Space-time budgets and activity studies 361

structure, for historical, institutional, and other reasons, is rarely a direct response to
the types of activity patterns currently displayed by urban residents. Nor, unless we
disregard significant political and economic forces, can it be argued that urban
structures are 'self-adjusting' and in the long run achieve 'equilibrium' with respect to
the activity patterns and preferences of urban households. As Keynes said, "In the
long run we are all dead"; and equilibrium theories (see Robinson, 1964, pp.65-80)
have severe limitations where locational changes through time are concerned,
particularly so if we agree with Harvey (1971) that modern cities are in a continuous
condition of 'differential ^equilibrium' (see below). Making urban structure more
responsive to the activity patterns of urban residents should indeed be one of the
fundamental and continuing tasks of planning. Therefore, rather than emphasize
their formative influence on spatial structure, it may be more useful to study how
activity patterns are themselves shaped by existing environments. According to
Meier (1962), time-budgets will not 'make or break' existing urban theories, but they
are likely to be increasingly useful in helping "to redress certain market value biases"
in urban decision-making.

7 Constraints and choices

Accessibility is a key factor in determining both the extent to which behaviour is
shaped by spatial environment, and the way in which households evaluate residential
locations. Accessibility, however, is a very complex set of variables. Their 'objective'
values (in time or money) vary widely with different population groups, transport
modes, and environments, and their implications also differ greatly depending on a
wide range of factors (such as age, income, activity commitments). The relevant set
of important accessibility opportunities varies both with life-cycle stage and with
'life style'. Chapin (1968) suggests that a relevant set may be inferred from time-budget
data; but this has severe limitations if some 'opportunities', which are felt to be relevant
by the respondents, are not revealed in activity patterns simply because they are too
inaccessible.
Social and spatial influences on behaviour are often highly interrelated and hard to
unravel. If (to reduce social variability) residence movers are studied, intended activity
changes should, ideally, be distinguished from unintended changes, and both types of
change should be related to the 'objective' structures and the perceived structures of
the new and old environments. However, in many housing studies, accessibilities are
given only partial or secondary consideration, or only one of the relevant accessibility
opportunities (for example, workplace) is considered. [Dwelling quality is probably
easier to conceptualise and may often be the primary factor, initially at least, in
intra-urban residence change—see Willmott (1967).] Activity patterns, too, are
usually treated in piecemeal fashion. Time-budget research suggests that the amount
of time which employed persons can spend at home on a working day decreases
as city size increases (Chombart de Lauwe, 1960; Hagerstrand, 1970b)—a
generalization with interesting implications for 'optimum' urban size and structure.
However, very little is known about critical or 'threshold' accessibility levels
('objective' or 'perceived', minimum or maximum), and there is some confusion
about the relative importance of social and spatial factors and how they interrelate
with behaviour.
From studies of American skilled-worker and lower middle-class households who
moved out to suburban locations, Berger (1960) and Gans (1967) concluded that
change of residence produced few, mostly minor or intended, behaviour changes.
Berger was mainly concerned with showing that suburban home-owning workers do
not become 'middle-class' in their social behaviour, and it has been suggested (Darke and
Darke, 1969) that Gans underestimated the influence of spatial environment because of
362 J Anderson

over-reaction to 'determinism'. From a national (USA) survey of recent movers,

Stegman (1969) found that suburban respondents were more concerned with
'neighbourhood quality' than with accessibility to other parts of the metropolitan
region. Increased living space was a major consideration, while locational convenience
was among a variety of infrequently mentioned factors. However, local accessibilities
may be subsumed in 'neighbourhood quality'. The same survey showed that, in major
metropolitan regions, motorway developments and decentralization combine to make
basic urban opportunities more accessible to suburbanites than to residents of the
inner city. Suburbanites spent less time getting to selected locations including
'downtown' (closer to inner city residences in physical distance terms), and Stegman's
conclusions were that many suburban families "do not have to trade-off accessibility
for savings in location rent; they can have both". Conversely, accessibility versus
living-space formulations also have limited relevance for poor black households, who
have little choice but to live in the congested core of a large metropolis. For similar
though less disparaging conclusions on residential rent theory, see Wheeler (1970).
Urban sprawl and motorways have "interwoven the problems of affluence and
poverty" (Thompson, 1967). With the undermining of public transport services,
accessibilities have decreased for those without private transport. Thompson suggests
the riots in the Watts ghetto, Los Angeles, were due in part to "the growing
inaccessibility of decentralizing job opportunities". Such urban structural changes are
seen by Harvey (1971) as inequality generating mechanisms which reflect inequalities
of political power and control over urban processes. Their effect is to redistribute
'real incomes' more inequitably in the population. The more advantaged sections of
urban society usually do not take so long to adapt to environmental changes, and/or
they generally have more power to direct these changes to their own relative advantage.
Hence, in an 'environment' of technological and other changes, urban systems continue
in a state of 'differential disequilibrium'. Real income depends partly on accessibilities
(including proximity to disamenities), and maximum and minimum accessibility-
thresholds merit more explicit attention in planning which aims to reverse inequality-
generating trends in spatial development.
Stegman's findings (above) do not apply to smaller American cities without radial
motorways, nor would similar findings generally be expected for British cities; and
here inequality generating spatial trends take different and usually less pronounced
forms. Accessibility decreases do emerge as significant in studies of tenants who are
relocated in suburban council estates (Willmott, 1967; Darke and Darke, 1969).
Journeys to work generally cost more, take longer, and more time is spent away from
the home neighbourhood; 'extended families' may break down; local job
opportunities for women are lacking; and both 'home-centredness' and juvenile
delinquency may be increased by the scarcity or non-existence of facilities within
easy reach of home. Such location-induced problems result only occasionally in
households returning to the inner city (Willmott, 1967); but this choice is extremely
limited by inner-area housing deficiencies. It is further constrained, in London at
least, by increasing competition from wealthier households choosing centrally located
housing (for example, their 'recolonizations' in Islington, Victoria and the Barbican
settlement)—they too find locational disadvantages in suburban living, although the
disadvantages are not necessarily of the same degree.
Spatial behavio.ur (home locating as well as daily activity patterns) can be visualized
as the outcome of choices which reflect people's motivations and values (Chapin and
Logan, 1969). However, such choices are realized within the constraints set by
personal circumstances, accessibilities, and environment, and the environment can be
seen as "a highly institutionalized power- and activity-system" (Hagerstrand, 1970a).
The limits of free choice are generally wider for the more wealthy and spatially
Space-time budgets and activity studies 363

mobile, and narrower for the poor, the old, those with young children, those without
a car, and so on. The 'options' in many situations are few (especially for the latter
group) or they may be reduced to 'Hobson's choice'.
Constraints are implicit in 'choice', but, depending on the relative emphasis given to
'positive' and 'negative' determinants (see Section 4 above), we may make a crude
distinction between 'choice' and 'constraint' oriented approaches to behaviour (Pahl,
1970). If psychological motivations of spatial behaviour (such as the 'rules' of
spatial choice, see Section 6 above) are sought, observed behaviour must be seen in
terms of choices between alternative courses of action. However, there is the danger
that important constraints may be underestimated or ignored, at least in the more
extreme forms of 'choice' approach. Observed behaviour can be thought of as
reflecting the actors' values and motivations, but it also reflects the constraints of
environment and personal circumstances. If highly constrained situations are not
recognized as such, observed behaviour may be interpreted misleadingly as what
people 'choose' rather than 'are forced' to do. With the prevailing emphasis on
'choice', situations where actors do not have effective choice may be neglected (for
example, the public housing sector in residential location and mobility studies, and
obligatory as opposed to leisure activities in time-budget studies); or, in the absence
of effective choice, people may be asked to decide between hypothetical alternatives
which in reality they have little hope of achieving.
The less common preoccupation with 'negative' determinants is exemplified in the
similar constraint orientations adopted by Pahl (1970) and Hagerstrand (1970a). In
attempting to redefine urban sociology, Pahl advocates studying "the pattern of
spatial and social constraints which operates differentially in given localities" and
which fundamentally affects people's 'life chances'—a framework with implications
for at least a partial re focusing of social geography. His argument is basically that
there are fundamental spatial (time, cost, distance) and social constraints on access to
scarce urban resources and facilities. The social constraints reflect the distribution of
wealth and power in society, they include bureaucratic rules and procedures, and the
actions of 'social gatekeepers' (such as local government officials, landlords,
employers) who help regulate the quality and accessibility of opportunities (for
example, education, employment, housing). Conflicts of interest in this socio-spatial
system are inevitable, and the greater the scarcity of valued opportunities the greater
the conflict. Pahl sees populations limited in their access to facilities as 'dependant
variables', while the 'gatekeepers' are the 'independant variables'. Noting that the
current emphasis on 'diversity and choice' in physical planning implies that access to
facilities is an 'independant' variable, in contrast to being, in his opinion, dependant
on allocation by 'gatekeepers', he suggests that there are ideological differences
underlying the variation between choice and constraint orientations (for example,
differing attitudes to 'free market' mechanisms).
The interrelationships between space and time uses have perhaps been worked out
most clearly by Hagerstrand and his colleagues (Carlstein et al, 1968). Influenced by
the 'physicalists', they focus on what Hagerstrand calls "the space-time mechanics of
constraints". Three interacting groups are identified: (1) capability constraints
include physiological regulators (such as sleep and age), and ability to move around in
physical space; (2) coupling constraints, operating within the first set of constraints,
include activity commitments, and the need for timing and synchronizing activities in
space; and (3) authority constraints refer to limitations and control of access—they
occur at different levels to produce hierarchies of 'space-time domains' (for example,
immigration control, domicile requirements, legal shop opening hours, and parking
restrictions). Particular space-time environments in the form of detailed computer
models, are evaluated on this conceptual basis (Hagerstrand, 1970a, 1970b) for the
364 J Anderson

degrees of constraint they impose on hypothetical activity schedules. These are

realistic because they are based on empirical surveys (Lenntorp, 1970; Martensson,
1970). Subjective perceptions of these 'objective' environments (for example,
inadequate or inaccurate information) are not considered, though in principle they
could be specified as additional constraints in the given activity schedule inputs.
In the 'constraints' approach, the choices decided on within the limits of the
constraints are not considered explicitly, and, as the psychological motivations of
behaviour decisions are not uncovered, it has limitations in predicting individual
behaviour. Until it has been applied in a wide variety of environments this approach
may tend towards the idiographic, but it is highly relevant to the planning of spatio-
temporal organization in particular areas. Nor should the crude distinction between
'choice' and 'constraint' orientations be overstressed. For instance, Maw (1969)
constructed a leisure model in the belief that constraints on 'free time' activity are
"so limiting that it seems possible to predict usefully what people will do in a given
situation"; to demonstrate its potential value in planning and time-tabling the provision
of facilities, he uses the model to predict usage of a municipal swimming pool.
Constraints on 'free time' include "a relatively fixed framework of essential activities"
(for example, primary work, shopping, and attendant travel) and their associated time
and place ties. In addition, there are 'discretionary' activities with varying degrees of
commitment, and locational ties in space-time (such as overtime, house repairs, and
optional shopping). Maw emphasizes that the concept of freedom in leisure behaviour
should be sufficiently broad to include activity commitments, monetary costs (see
also, Milstein, 1961), social exclusiveness, and required physical effort. These define
the limits of 'free-choice'; actual choice then depends on additional personal
preference factors.

8 Conclusions
Space-time budgets provide a comprehensive description of activity routines which are
not directly observable because of their spatial and temporal extent. Although the
recorded period is usually of necessity short, fundamental daily and weekly activity
rhythms can be studied, and people's identity over time (during the recorded period)
is preserved. The organization of their activity patterns can be treated as a 'system',
with explicit recognition of the fact that particular activities effect and constrain each
other, while the institutions to which people belong (such as households) and with
which they interact (shops, schools, recreation places) can be treated in a similar way.
Thus a particular type of activity, for example, shopping, might be analysed in the
context of the overall space and time uses of the shoppers and their households in
relation to shop locations and business-hours.
A set of activity records provides a wealth of data which may be used for a variety
of purposes. They are expensive to collect, however, and because of this, collection
might best be undertaken by public authorities and large planning agencies (as, for
example, in Hungary). They are also difficult to analyse adequately. In particular,
obtaining adequate summary measures of a person's overall space-time behaviour
remains a major problem. Faced with a wealth of information on diverse activities,
their attributes, and space-time locations, the usual procedure has been to abstract
from the data-set, collapse the temporal or the spatial dimensions, and carry out one
or more partial analyses. In this respect, Hagerstrand's computer modelling of space-
time environments is an outstanding exception. The analysis of space-time uses,
albeit partial analysis, in order to be most useful, should ideally be linked (via
questionnaire and other data) to the study of less tangible 'covert' behaviour (for
example, environment perception, attitudes, motives), and/or to the analysis of
environmental constraints (socio-economic and spatial accessibilities). Studies of the
Space-time budgets and activity studies 365

space-time budgets of households might usefully be complemented by studies of the

temporal organization and spatial linkages of selected institutions and public places
(such as workplaces, schools, social service places and public libraries) which have a
direct bearing on the lives of household members.
Chapin (1968) notes that most concepts of urban spatial structure are built around
land rent theory and market mechanism 'organizers', and he questions the wisdom of
relying so heavily on one system of thought, even if alternatives seem incomplete by
comparison. Hagerstrand (1970a) claims that: "Nothing truly general can be said
about aggregate regularities until it has been made clear how far they remain invariant
for organizational differences at the micro-level." The findings of Stegman (1969)
lend weight to these views. His findings, and others in similar vein (see Section 7),
do not invalidate the land rent approach, but they suggest that for it to be
appropriate urban space has to be transformed in a very complex manner (for
example, into a time-surface map on which distance corresponds not to mileage but
to travel time—see, Angel and Hyman, 1970). It is apparent that alternative
approaches to urban analysis can have at least equal validity, and that alternatives,
even if incomplete, are in fact badly needed.
Space-time budgets offer little hope in the quest for 'independant' behavioural
postulates, from which improved theories of spatial structure could be derived, because
of the strong influence of environmental constraints on observed behaviour. A more
useful task, as far as urban planning is concerned, is analysing how spatial environments
effect the activity patterns of different types of people and household (see, Chombart
de Lauwe, 1963). Such studies could provide a basis for planning and modifying
environments so that less constraint is imposed on desired behaviour patterns.
Account must of course be taken of the wider economic and political forces in
operation. Micro-level activity studies (and studies in 'behavioural geography' in
general) run the risk of losing sight of the social contexts within which the observed
people are 'behaving'. While space-time budgets may "express value in terms relevant
to social organization rather than the units of the market place" (Meier, 1962), economic
factors are among the most important determinants of human behaviour.
References
Adams, J., 1968, "A review of behaviour and location: foundations for a geographic and dynamic
location theory, Part 1, by A. Pred", London School of Economics and Political Science,
Graduate Discussion Paper 20, London.
Anderson, J., 1970, "Time-budgets and human geography: notes and references", London School
of Economics and Political Science, Graduate Discussion Paper 36, London.
Anderson, J., 1971, "Living in urban space -time", Architectural Design, 41, 41-44.
Anderson, J., Goddard, J., 1969, "Some current approaches to human geography in Sweden",
London School of Economics and Political Science, Graduate Discussion Paper 33, London.
Anderson, N., 1961, Work and Leisure (The Free Press, New York).
Angel, S., Hyman, G., 1970, "Urban velocity fields", Environment and Planning, 2, 211-224.
Arias, F., Diaz, A., 1970, "Golborne Ward: observation and time-budget analysis of public places",
Planning Department, Architectural Association, School of Architecture, London, mimeo.
Baker, A. M., 1969, "Models of spatial behaviour in urban environments", paper presented at
I.G.U. Commission on Quantitative Methods, Ann Arbor, August; Research Paper 15, Centre for
Urban and Community Studies, University of Toronto, Toronto, Canada.
Becker, G. S., 1965, "A theory of the allocation of time", Economic Journal, 75, 493-517.
Berger, B. M., 1960, Working Class Suburb (University of California Press, Berkeley).
Blight, G. M., 1970, "Urban areas and the analysis of space and time use", Geography Department,
London School of Economics, mimeo.
Brail, R. K., 1969, Activity System Investigations: Strategy for Model Design, Ph. D. Thesis,
Planning Department, University of North Carolina, Chapel Hill.
Brown, L. A., Horton, F. E., 1970, "Functional distance: an operational approach", Geographical
Analysis, 2, 76-83.
Burenstam Linder, S., 1970, The Harried Leisure Class (Columbia University Press, New York).
366 J Anderson

Burton, T. L., Noad, P. A., 1968, Recreation Research Methods: A Review of Recent Studies,
Centre for Urban and Regional Studies, Occasional Paper 3, University of Birmingham, England.
Calhoun, J. B., 1963, "Population density and social pathology", in The Urban Condition,
Ed. L. Duhl (Basic Books, New York).
Carlstein, T., Lenntorp, B., Martensson, S., 1968, "Individers dygnsbanor i nagra hushallstyper",
Urbaniserings Processen, 17, Department of Geography, University of Lund, Lund, Sweden.
Chapin, F. S., 1965, Urban Land Use Planning (University of Illinois Press, Urbana).
Chapin, F. S., 1968, "Activity systems and urban structure: a working schema", Journal of the
American Institute of Planners, 34, 11-18.
Chapin, F. S., Hightower, H. C , 1965, "Household activity patterns and land use", Journal of the
American Institute of Planners, 31, 222-231.
Chapin, F. S., Hightower, H. C , 1966, "Household activity systems—a pilot investigation"
(University of North Carolina, Chapel Hill).
Chapin, F. S., Logan, T. H., 1969, "Patterns of time and space use", in The Quality of the Urban
Environment, Ed. H. S. Perloff (Resources for the Future, Inc., Washington, D.C.).
Chombart de Lauwe, P. H., 1960, Famille et Habitation, Volume 2, Centre National de la Recherche
Scientifique, Paris.
Chombart de Lauwe, P. H., 1963, "The social sciences, urbanism and planning", International
Journal of Comparative Sociology, 4, 19-30.
Clarke, W. A. V., 1968, "Consumer travel patterns and the concept of range", Annals of the
Association of American Geographers, 58, 386-396.
Converse, P. E., 1968, "Time budgets", in International Encylopedia of Social Science (Macmillan,
New York).
Cox, K. R., Golledge, R. G. (Eds.), 1969, "Behavioural models in geography", in Behavioural
Problems in Geography, Studies in Geography, 17, pp.1-13, Northwestern University.
Cullen, I., 1970, "Algorithmic approaches to the recognition of activity patterns", Seminar Paper 19,
Joint Unit for Planning Research, University of London.
Curry, L., 1967a, "Quantitative geography, 1967", Canadian Geographer, 11, 265-279.
Curry, L., 1967b, "Central places in the random spatial economy", Journal of Regional Science,
7 (Supplement), 217-238.
Darke, R., Darke, Jane, 1969, "Suburban housing estates: social composition and characteristics",
CES-WP-40, Centre for Environmental Studies, London.
Doherty, J. M., 1969, "Developments in behavioural geography", London School of Economics and
Political Science, Graduate Discussion Paper 35, London.
Engels, F., 1892, The Condition of the Working-class in England (1844) (Panther, London).
Eyles, J., 1970, "Space, territory and conflict", Geographical Paper 1, University of Reading,
Reading, England.
Firth, R. (Ed.), 1967, "Themes in economic anthropology", Association of Social Anthropologists
Monograph, London.
Foster, C. D., Beesley, M. E., 1963, "Estimating the social benefit of constructing an underground
railway in London", Journal of the Royal Statistical Society, Series A, 126, 46-92.
Gans, H. J., 1967, The Levittowners: Anatomy of Suburbia (Pantheon, New York).
Gould, P. R., 1969, "Methodological developments since the fifties", in Progress in Geography, 1,
Eds. C. Board, R. J. Chorley, P. Haggett, D. R. Stoddart (Edward Arnold, London).
Gurvich, G., 1964, The Spectrum of Social Time (Reidel, Dordrecht).
Hagerstrand, T., 1968, "Methods and new techniques in current urban and regional research in
Sweden", Plan, 22, 3 - 1 1 .
Hagerstrand, T., 1970a, "What about people in regional science?", Presidential Address European
Regional Science Association, Copenhagen, August; Papers of the Regional Science Association,
24,7-21.
Hagerstrand, T., 1970b, "Tidsanvandning och omgivningsstrukture", in Urbaniseringen i Sverige:
en geografish samhallsanalys, Stockholm.
Haggett, P., 1968, "The spatial structure of city regions", Social Science Research Council/Centre
for Environmental Studies Joint Conference, "The future of the city region", CES-WP-6,
Volume 2, Centre for Environmental Studies, London.
Harvey, D., 1969, Explanation in Geography (Edward Arnold, London).
Harvey, D., 1971, "Social processes, spatial form and the redistribution of reaLincome in an urban
system", in Proceedings, 22nd Colston Research Society Symposium, Eds. M.Chisholm, A. Frey,
P. Haggett (Butterworths, London).
Hemmens, G. C , 1966, "The structure of urban activity linkages" (University of North Carolina,
Chapel Hill).
Space-time budgets and activity studies 367

Hurst, M. E. E., 1969, "The structure of movement and household travel behaviour", Urban Studies,
6, 70-82.
King, L. J., 1969, "The analysis of spatial form and its relation to geographic theory", Annals of
the Association of American Geographers, 59, 573-595.
Lenntorp, B., 1970, "PESASP—en modell for berakning av alternativa banor", Urbaniserings
Processen, 38, Department of Geography, University of Lund, Lund, Sweden.
Lewis, O., 1959, Five Families: Mexican, Case Studies in the Culture of Poverty (Basic Books, New
York).
London Traffic Survey, 1964, 1966, 2 volumes (Greater London Council, London).
Lundberg, G. A., Komarowsky, M., Mclnerny, Mary Alice, 1934, Leisure, A Suburban Study
(Columbia University Press, New York).
Marble, D. F., 1959, Section 3 of Studies of Highway Development and Geographic Change,
Ed. W. L. Garrison (University of Washington Press, Seattle).
Marble, D. F., 1967, "A theoretical exploration of individual travel behaviour", Quantitative
Geography, 1, Studies in Geography, 13, (Northwestern University), pp.33-53.
Marble, D. F., 1970, "Travel diaries—problems of data collection and analysis", Paper presented at
the Cambridge meeting, Institute of British Geographers, London.
Martensson, S., 1970, "Tidsgeografisk beskrivning av stationsstrukture", Urbaniserings Processen, 39,
Department of Geography, University of Lund, Lund, Sweden.
Maw, R., 1969, "Construction of a leisure model", Official Architecture and Planning, August,
924-935.
Mead, Margaret (Ed.), 1954, Cultural Patterns and Technical Change (UNESCO, New York).
Mead, W. R., 1958, "The seasonal round:- a study of adjustment on Finland's pioneer fringe",
Tijdschrift voor Economische en Sociale Geografie, 49, 157-161.
Meier, R. L., 1962, A Communications Theory of Urban Growth (MIT Press, Cambridge, Mass.).
Milstein, D. N., 1961, "An economic approach to leisure analysis", Social Problems, 9, 17-31.
Moore, E. G., 1970, "Some spatial properties of urban contact fields", Geographical Analysis, 2,
376-386.
Moore, W. E., 1963, Man, Time and Society (John Wiley, New York).
Morrill, R. L., Pitts, F. R., 1967, "Marriage, migration, and the mean information field: a study in
uniqueness and generality", Annals of the Association of American Geographers, 57, 401 -422.
Mumford, L., 1934, Technics and Civilization (Harcourt, Brace & Co., New York).
Neurath, O., 1944, "Foundations of the social sciences", International Encyclopedia of Unified
Science, 2, 1-51, University of Chicago, Chicago.
Nichols, Vida, 1969, "An institution in metropolis", Seminar Paper 12, Joint Unit for Planning
Research, University of London.
Nystuen, J. D., 1959, Section 4 of Studies of Highway Development and Geographic Change,
Ed. W. L. Garrison (University of Washington Press, Seattle).
Nystuen, J. D., 1963, "Identification of some fundamental spatial concepts", in Spatial Analysis,
Eds. B. J. L. Berry, D. F. Marble (Prentice-Hall, Englewood Cliffs, N.J.), pp.35-41.
Nystuen, J. D., 1967, "A theory and simulation of intra-urban travel", Quantitative Geography, 1,
Studies in Geography, 13, Northwestern University, Evanston, Illinois, pp.54-83.
Pahl, R. E., 1970, Whose City? And Other Essays on Sociology and Planning (Chapter 13)
(Longmans, Green, London).
Quarmby, D. A., 1967, "Choice of travel mode for the journey to work: some findings", Journal
of Transport Economics and Policy, 1, 273-314.
Robinson, Joan, 1964, Economic Philosophy (Penguin Books, Harmondsworth).
Rushton, G., 1969, "Analysis of spatial behaviour by revealed space preference", Annals of the
Association of American Geographers, 59, 391-400.
Sommer, R., 1969, Personal Space: The Behavioural Basis of Design (Prentice-Hall, Englewood
Cliffs, N.J.).
Sorokin, P. A., Berger, C. Q., 1939, Time-budgets of Human Behaviour (Harvard University Press,
Cambridge, Mass.).
Stegman, M. A., 1969, "Accessibility models and residential location", Journal of the American
Institute of Planners, 35, 22-29.
Szalai, A., 1966, "The multinational comparative time budget research project", American
Behavioural Scientist, 10, 1-4.
Thompson, W., 1967, "Urban economics", Chapter 4 of Taming Megalopolis, Volume 1,
Ed. H. W. Eldredge (Anchor Books, New York).
Webber, M. M., 1964, "Culture, territoriality and the elastic mile", Papers of the Regional Science
Association, 13, 59-69.
368 J Anderson

Wheeler, J. O., 1970, "Transport inputs and residential rent theory. An empirical analysis",
Geographical A nalysis, 2 , 4 3 - 5 4 .
Wilkins, M. B., 1968, "Biological clocks", Advancement of Science, 24, 273-283.
Willmott, P., 1967, "Social research and new communities", Journal of the American Institute of
Planners, 33, 387-398.
Wilson, N. L., 1955, "Space, time and individuals", Journal of Philosophy, 52, 589-598.
Wolpert, J., 1965, "Behavioural aspects of the decision to migrate", Papers of the Regional Science
Association, 15, 159-169.

p © 1971 a Pion publication printed In Great Britain