SOUTH AFRICAN URBANISATION DYNAMICS
A N D THE NORMALISATION THESIS
ALISON TODES
Since the 1980s, the argument that South African state policy should focus
on developing the metropolitan areas and other places demonstrating
significant comparative or competitive advantages has become an important
one. Born in the context of a critique of apartheid policy, it is also seen as a
way of correcting the distorted and inefficient settlement patterns created
under apartheid. A significant building block of this argument is that, with
the termination of apartheid controls, people would migrate to the cities,
and 'artificially created' s e t t l e m e n t s w o u l d wither away (UF, 1990:
Tomlinson, 1990; Roux, 1991). The characteristic a p a r t h e i d spatial
disjunctures between population and economic activity at a national and
r e g i o n a l scale w o u l d d i s a p p e a r as p e o p l e m o v e d closer to work,
'normalising' the settlement pattern.
Yet there are reasons to doubt the plausibility of the normalisation assumption. Analysis of recent literature on urbanisation and settlement suggests that while there is indeed significant movement to the cities, and to
other centres of economic growth, there are also movements into areas with
weak or declining economic bases (especially smaller towns). Further, movement out of many old resettlement areas, and places experiencing economic
restructuring or decline, has been limited. Therefore, while there is some
correction of apartheid spatial disjunctures, many such disjunctures remain
and in some cases, new forms of disarticulation may be emerging. These
dynamics suggest that past settlement patterns may not be as ephemeral
as suggested in the normalisation thesis.
On the face of it, these dynamics might seem to defy migration theory.
That people move for economic reasons is almost considered a truism in
migration theory. This article explores various theories that might help to
explain these patterns, and, in particular, to examine why people might
remain in places with weak or declining economic bases. The plausibility
of the normalisation thesis is explored further through examination of
the case of Newcastle, KwaZulu-Natal, a town of around 350 000, which
2
URBAN FOR UM
grew enormously under the combined policies of forced removals, influx
control, and industrial decentralisation. Although Newcastle is a large town,
it has experienced economic restructuring and has faced substantial employment losses in recent years. Yet the town has experienced net in-migration since the collapse of influx control, s e e m i n g l y d e f y i n g the
normalisation thesis.
The study thus raises questions about the normalisation thesis, and suggests that migration patterns are more complex than it assumes. Declining
levels of employment, crime and violence in cities are some of the factors
limiting the extent to which 'normalisation' is occurring. In this context,
the securities, assets and social networks people have built up in places are
not easily abandoned. This however, raises questions about a sole focus on
metropolitan areas and other places demonstrating comparative and competitive advantages.
"/.-'
i
I
I
17°E
22°E
REPUBLICOF SOUTHAFRICA
PROVINCES
f
"..
j
BOTSWANA
NORTHERN
]
F"\.....j"
,..
=o
r.,
IPretoria
,r',-JJohann"sburg
=i ~..-~" -_~
.o.,.w,,,
"
[
"~"
"
ff
B4oemfontein.t
~
~.
/~
,
NORTHERN CAPE
/i.Newcaste
"
~
~..~.~X.,.5.~.~,-,
ATLANTIC
OCEAN
q
"~
irlo~n
"
k\
CapeTov.,~
17°E
I
Bo.e,o,h
/
/
~:~"~ EASTERN CAPE
r
WESTERN
/
LESOTHO .i
17°$
..%,....,,T.i
yNO
~
~ /.....
°,~
I
/"d'~X.. KWAZULU
NATAL
•
k
,"
"7" O Bilanyoni
,~.~
dA
.e,sp,0~..~,
-'~"1 .... ,,..-"
~2~fffFREE STATE
..,....,...,~,,
~.
<.7,~,o.¢"
NAM I BIA
•
~'*"
,~.j~.t"
" " ' ~ _.,_...
2¢s
,00
i
--~ooOon
CAPE
INDIAN
OCEAN
22°E
I
27°E
I
Figure 1: Provinces and towns in South Africa, and the Newcastle area
SO U T H A F R I C A N U R B A N I S A T I O N D Y N A M I C S
3
S O U T H A F R I C A N U R B A N t S A T t O N D Y N A M I C S I N THE 1990S 1
The following discussion is based on a review of secondary literature on
urbanisation and settlement patterns in the 1990s, due to difficulties in
using census data for analysis 2. The literature provides uneven coverage of
the issues. While CDE (1995, 1996, 1998) and UF (1994) provide useful overviews and national coverage by settlement type, demographics are based
either on projections of the 1991 census, or use of town estimates, also
linked to the 1991 census. Other studies are largely regional or local in
their focus, with particular concentration on KwaZulu-NataP, the Eastern
Cape and Western Cape 4, and the Free State s (Figure 1). A few studies provide conceptual overviews and summaries (Wolpe, 1997; Mabin, 1997),
sometimes focused on a particular settlement type (Dewar, 1996). The literature provides a good overview of dynamics, particularly around small
towns and resettlement areas, but trends are less clear for secondary cities
in the 1990s.
Metropolitan areas
In the late 1970s and early 1980s, rapid migration occurred into most
metropolitan areas, leading to the collapse of apartheid influx controls, and
reversing previous trends. Estimates by the Urban Foundation (UF, 1989)
at the time projected continuing rapid movement into these areas, but by
the late 1980s and early 1990s, it was apparent that flows into most
metropolitan areas had not been sustained. Studies in the mid-1990s suggest
a more varied picture. Rapid in-migration does appear to be occurring in
some cities, and not in others, reflecting both conditions in the sending
areas, and in the cities themselves.
Cape Town (Western Cape) has experienced rapid i n - m i g r a t i o n as
a consequence of flows from the Eastern Cape, where sharply declining
conditions in rural e c o n o m i e s from 1986 has led to rapid m o v e m e n t
both within the province, and to cities outside of it (Bank, 1997, Cross et al,
1999). G a u t e n g has also been a major recipient of these m o v e m e n t s
(Bank, 1997). Movement into the Western Cape and to Cape Town in part
reflects their better relative economic performance in the 1990s, as well
as the huge differentials in incomes and employment between these areas
and the Eastern Cape (Cross et al, 1999). The existence of well-established
migration routes between the Eastern and Western Cape is also significant.
Gauteng also seems to have been a major focus for migration in the 1990s,
as it has been historically. Its growth has been estimated at 2,5 percent p.a.
between 1991 and 1995, compared to 2,1 percent p.a. nationally (Tomlinson,
4
URBAN FORUM
1997). Tomlinson (1997) suggests that while this in-migration occurred following a period of severe rationalisation of employment, when more jobs
were shed in Gauteng than elsewhere, levels of unemployment are far lower,
and incomes higher than in other areas, and high rates of in-migration can
be expected to continue.
Most studies of Durban in the 1990s suggest that migration into the city
has slowed considerably since the mid 1980s (Hindson and McCarthy, 1994;
Urban Strategy, 1995). There is little evidence to suggest that this has
changed in the 1990s. Its economic performance in the 1990s appears to
have been weaker than in previous decades, with significant job losses in
some key manufacturing sectors (see Morris et al, 1998;Valodia, 1998), and
a somewhat flailing tourism industry. Unemployment rates are far higher,
and income levels lower than in Gauteng and the Western Cape. Cross et
al (1998) suggest that rising unemployment and declining real incomes are
limiting in-migration, especially from rural areas. Further, urban jobs are
increasingly occupied by established residents who have the skills to acquire them.
An important pattern however has been a movement into rural areas on
the edges of D u r b a n - - b o t h from other parts of the province, and from
Durban itself. These areas offer lower service costs, and the opportunity for
livelihoods derived from the natural resource base, as well as from the city
(Cross et al, 1996, 1998). Similar patterns are occurring around other towns in
KwaZulu-Natal as well. Areas of this sort are increasingly becoming the focus
of 'back migration' , when movement to cities fails, but are sometimes preferred by the better-off (Cross et al, 1998). Although the particular pattern of
land tenure around KwaZulu-Natal urban areas enables this process, Cross
et al (1996, 1999) suggest that similar patterns based on multiple livelihoods partially drawing on a natural resource base might also be occurring
on the edges of Cape Town, and around smaller towns elsewhere.
Less work is available on the smaller metropoles, but Bank's (1997) study
of East London (Eastern Cape) is illuminating, and suggests some different
trends to those noted in other cities. Bank (1997) shows a rapid movement
into East London in the late 1980s and early 1990s, although migration
now appears to be slowing as the local economy stagnates. Migration to
East London was linked to processes displacing people from both tribal
areas and commercial farms in the Eastern Cape (discussed more fully below), which also underpinned strong movement to small towns. Interestingly, migration into East London is largely from rural areas nearby: it does
not appear to operate as a major target for regional migration, or as part of
step migration. Migration largely occurs on the basis of existing networks,
and patterns of circular migration persist. Bank (1997) makes a strong plea
for u n d e r s t a n d i n g the local dynamics of u r b a n i s a t i o n and avoiding
homogenising assumptions.
SO UTH AFRICAN URBANISATION DYNAMICS
5
Large Towns
Large towns between 50 000 and 500 000 grew most rapidly in the 1980s
(CDE, 1996). Many of these towns grew as a consequence of industrial
decentralisation, or their role in homeland administration, both of which
were part of apartheid policies to contain African people within independent or self-governing 'homelands' . Displacement from surrounding rural
areas has also played a role in the growth of these areas. Patterns in the
1990s are not clear: whilst displacement from rural areas has continued,
many of these towns are experiencing economic restructuring and some
may be declining in economic terms. As UF (1994) notes, these towns are
often vulnerable in a context of restructuring since they frequently rely on
a limited number of sectors. The decline of mining, the restructuring of
particular sectors within manufacturing as South Africa opened up to the
global economy, the ending of the old era of industrial incentives, challenges to low waged work on the periphery as a consequence of growing
unionisation and international competition (Hart and Todes, 1997), and the
rationalisation of administration may all be affecting these towns. Patterns
vary between places, however, and some such towns may be experiencing
economic growth as a consequence of their role as regional service sectors,
or due to growth in industry or tourism (eg. Richards Bay and Nelspruit). It
is not clear how population movements in general have responded to these
changes, and patterns are likely to vary. Certainly, there is a rapid movement to places experiencing growth, such as Richards Bay and Knysna, but
there is not necessarily out-migration from places experiencing decline, as
the Newcastle study, discussed below, demonstrates.
Small Towns
CDE' s (1996) study of small towns suggested that there was considerable variation in patterns of economic and demographic growth in the 1990s.
Using figures for magisterial districts as a proxy, some 69 percent declined
in terms of economic output between 1991 and 1994. The economic position of small towns is usually strongly linked to agriculture in their hinterland, and their role within it, but some towns are growing due to their
roles in tourism or as retirement areas, their location relative to particular
patterns of accessibility, their role as service centres in densely populated
areas, and their ability to take on more diverse roles in some circumstances
(particularly close to metropolitan areas) (Dewar, 1996; CDE, 1996). In
m a n y commercial farming areas, small towns have been experiencing
long term decline due to concentration and centralisation within agriculture and processing industries, and as a consequence of depopulation of
farms.
6
URBAN FORUM
Patterns of population growth are also variable, and do not necessarily
correspond with economic growth. CDE's (1996) sample of population
growth trends in 30 towns between 1991 and 1995 suggests that many declining towns are also experiencing population growth, but about half of
all towns experienced net out-migration.
Studies of particularly the Eastern Cape (Bank, 1997, Cross et al, 1999)
and the Free State (Krige, 1999) suggest a massive movement into small
towns since the mid-1980s. Krige (1995) argues that some 250 000 people
moved from farms to small towns in the Free State between 1985 and 1995,
and Bank (1997) notes that 43 percent of new squatters in small towns in
the Eastern Cape in 1993 were from commercial farms. Similar movements
are occurring in KwaZulu-Natal (also to dense rural settlements around
towns--see Cross et al, 1998), but not necessarily of the same order. A number of processes underpin these patterns. Firstly, there has been significant
displacement from commercial farms as a consequence of drought, declining material c o n d i t i o n s on farms, m e c h a n i s a t i o n , fears of g r o w i n g
unionisation and the imposition of labour legislation (Bank, 1997), and fears
of land claims (Marcus et al, 1994). New rounds of forced removals are sometimes occurring, and farmworkers are emerging as an important part of
housing demand in small towns (in KwaZulu-Natal). Studies suggest that
farmworkers tend to move to the nearest town, rather than further afield.
Population displacement is also occurring from within tribal areas to small
towns in ex-homeland areas. Patterns are most pronounced in parts of the
Eastern Cape. Butterworth, for example, has grown extremely rapidly (at
over 10% pa in the 1990s: see Bank, 1997) despite the collapse of industrial
employment in the town. Similarly, growth has been dramatic in other towns
with weak, limited or declining economic bases, placing severe strain on
infrastructure and services (Bank, 1997; Cross et aI, 1999). Rates collection
is poor and local m a n a g e m e n t capacity is limited. Movement into these
areas occurred in a context where homeland governments and their control of land was collapsing, leading to land invasions and the development
of shack settlements. At the same time, conditions in rural areas were becoming more difficult: chiefs denied youth access to land as political tens i o n s rose; d r o u g h t a n d u n e m p l o y m e n t u n d e r p i n n e d m o v e m e n t
(particularly by w o m e n - h e a d e d households) in search of informal economic
opportunity and land (Bank, 1997). Movements are often to towns close by,
enabling links to home areas to be maintained (Bank, 1997, Cross et al,
1999).
In the KwaZulu-Natal context, Hindson and McCarthy (1994), and Kiepel
(1994) suggest that people may increasingly be moving closer to urban areas, but many of these places are also marginal or declining small towns,
raising questions about the extent to which these and similar patterns noted
above really constitute a shift away from the disjunctures created by apart-
SO UTH AFRICAN URBANISATION DYNAMICS
7
heid. Rather, new disjunctures between population settlement and economic
activity may be occurring.
The decline of migrant labour, and particularly, major retrenchments by
traditional employers of migrant labourers (such as the mining industry)
appear to be accelerating urbanisation as workers move closer to sources of
employment (Davies and Head, 1995), but in a number of cases these workers
are moving to towns close to mines, where employment is declining. Wives
are sometimes being abandoned in these processes, and may move on to
small towns as remittances decline (Bank, 1997; Davies and Head) 1995).
Resettlement Areas
In contrast to the assumption that resettlement areas 6 or 'closer settlements' are simply 'Apartheid' s dumping grounds', which would simply
disappear with the ending of Apartheid, recent research shows how many
of these places have persisted (Mabin, 1997) and are unlikely to whither
away. Surveys suggest that many such places are now preferred as residential locations by large numbers of people living there (Meth, 1998; CDE,
1998). Nevertheless perceptions are not uniform, and some would move
out if they could. Out-migration was evident in a number of such settlements from the late 1980s (see eg. Krige, 1996 on Botshabelo some 60kin
from Bloemfontein), and CDE (1998) argues that those who would like to
move out have already done so.
In some cases, people are moving into these settlements: the net outcome appears to be slow growth, but there are significant variations bet w e e n settlements. Mabin (1997) argues that the better located closer
s e t t l e m e n t s - - n e a r to transport routes and town peripheries--are growing
rapidly, and that new concentrated informal settlements are emerging in
similar places. Similarly Kiepel (1994), and Hindson and McCarthy (1994)
on KwaZulu-Natal suggest that informal settlements (including all dense
rural settlements) are concentrating around main transport routes and near
to towns, and that there has been a movement to these areas from more
distant parts of districts (see also Cross et al, 1996). Kiepel (1994) argues
that the smaller settlements (under 1000) have tended to disappear. Nevertheless, about a third are more than 15km from a town, and located within
marginal regions.
Resettlement areas themselves include a wide variety of places, some of
which are now fully incorporated within urban areas. Others, however, are
concentrated settlements within rural areas, some distance from towns and
cities, although they might legally be incorporated within them. Sizes vary
considerably, but large settlements can have several hundred thousand resid e n t s - f a r bigger than many towns. Some of the largest settlements are within
long-distance commuting range of cities, particularly around Gauteng.
8
URBAN FORUM
Many of these places consolidated over the years and even attracted
people who were not forced to move there, as they were easy to move into,
infrastructure and services were often better than in surrounding areas,
while costs were lower than in the cities. While some researchers have argued that these places are entirely dependent on subsidies and could disappear without them (Tomlinson and Krige, 1997), others suggest that the
role of subsidies is overstated. CDE (1998) found that the bus subsidy was
less significant than initially assumed as there had been a significant transfer
to taxis in many areas (although the existence of subsidised buses kept costs
lower). The CDE (1998) studies were focused on places where bus subsidies were a concern, but not all settlements are tied into commuting or are
dependent on bus subsidies (see Meth' s (1998) study of Bilanyoni).
Although many of these areas have a harsh history, incomes are generally better than in rural areas (but less than urban incomes), and infrastructure and services more fully developed. CDE (1998) suggests that given
lower residential costs, savings may be higher than the norm for Africans.
Households have invested significantly in housing, and have larger houses
and sites than could be afforded under current housing policy. Households
also say they prefer these areas due to lower crime rates and family networks, which are now well established.
CDE (1998) argues that unemployment levels are similar to cities, but
that economic dependency rates are high. A number of these places were
poorly performing industrial decentralisation points, and few jobs have been
created in the 1990s--and many lost. CDE (1998) argues that these areas
have largely depended on migrant or more often commuter employment,
but this may neglect the extent of local economic activity particularly in
the larger and more isolated settlements. Mabin (1997) for example talks
of 'vibrant commercial growth' and ' d e e p e n i n g layers of internal economic relationships' in some of these areas (p. 6).
It is evident that some of these patterns conform to the expectations of
the normalisation thesis, while others do not: there is both rapid movement to areas of growth, and movement into or stasis within some places
with weak or declining economic bases. The following section reviews migration theory, in order to derive explanations for those patterns that do
not conform to the normalisation thesis.
MIGRATION THEORY AND THE NORMALISATION THESIS
Although migration theory is largely oriented to explaining why people
move, it is also useful in analysing why movement may be more limited
than expected. Neo-classical theories of migration have tended to dominate
SOUTH AFRICAN URBANISATION DYNAMICS
9
the field, although other theories have presented important critiques of
these models, and new perspectives on migration.
Neo-Classical Theories
Common to most neo-classical models is the focus on individuals as key
decision-makes in migration, and the argument that migration is a response
to differentials in conditions between places. Neo-classical theories stress
the impetus which differences in income, wages, and levels of employment
give to migration, suggesting that migration is likely to result in spatial
equilibrium in the context of uneven development. In short, migrants are
likely to move out of economically stagnant or declining areas, to ones where
greater o p p o r t u n i t i e s are available. The d o m i n a n t perspective views
economic determinants as primary, although behavioural models have, to
some extent, stressed a wider range of factors at play. The influential
behavioural model, as developed by Wolpert (1965) for example, saw the
decision to migrate as a result of differences in the utility associated with
places. 'Utility' m e a s u r e d the i n d i v i d u a l ' s level of satisfaction (or
dissatisfaction) with a place, and the model predicted that where it was out
of line with their aspirations, they would move.
Although economic factors are seen as dominant, the selectivity of migration is recognised--in particular, education levels, skill levels, age and
income are considered (Todaro, 1981). Even within this perspective, then,
the probability of low-skilled unemployed workers migrating would be
lower than for more educated and skilled workers or professionals. Factors
that affect the economic calculus of migration, such as the costs of moving,
are also c o n s i d e r e d to s o m e extent. In the South African context,
Wittenburg's (1999) work suggests an increasing polarisation between urban and rural areas on the basis of skills, while Cross et al' s (1998, 1999)
research similarly suggests that access to urban labour markets for the unskilled rural poor is becoming more difficult, and that rural and urban areas may be becoming increasingly divided.
Studies within the neo-classical school have also demonstrated the way
in which government policies influence the spatial distribution of opportunities (Charney, 1993). Of importance here is the availability of social
security benefits, which dampen the need to migrate in the context of
unemployment. The existence of pensions as a basic form of livelihood for
many marginalised South African households could thus limit the propensity to migrate from areas with weak economic bases. Similarly, access to
subsidised (often almost free) services within settlements falling under
homeland rule in the past would have distorted migration patterns, but it
is not clear that subsidies were necessarily higher on the periphery than in
homeland areas within cities such as Durban or East London. The shift to a
10
URBAN FORUM
more systematic form of subsidisation of poor households through the Equitable Share of National Revenue for Local Government, coupled with a
greater emphasis on cost recovery, could however change the cost equation, and might lead to out-migration from some resettlement areas if it
does (see Meth, 2000).
Policies that influence spatial access to housing, education and health
might also be influential in distorting the workings of the market, although
in terms of the neo-classical perspective, they would be less influential than
policies which affected the spatial distribution of economic opportunities.
Years of providing housing in homeland areas as opposed to the cities have
influenced migration patterns, and in part underpin current ties to place,
as the case of Newcastle (below) suggests. CDE (1998) also cites better
housing in resettlement areas than is currently available in the cities as a
factor in households' choice to remain there. According to Mokate et al
(1999), government infrastructural expenditures since 1994 have tended to
favour urban areas (with the exception of water), although there has been
a recent shift towards rural areas. Past state policies of i n d u s t r i a l
decentralisation, the creation of homeland governments, and the development of public services (including a major tertiary education sector), were
instrumental in generating employment on the periphery (Liebbrandt and
Woolard, 1996), and in part u n d e r p i n n e d s e t t l e m e n t p a t t e r n s . The
rationalisation of the bureaucracy, the decline of homeland universities,
and the collapse of some industrial decentralisation points could however
alter these patterns.
Although the expectation within neo-classical theory is that the free
operation of the market will lead to spatial equilibrium, Milne (1993) suggests that the business cycle itself results in variations in levels of migration. In recessionary periods, migration slows, since the probability of
finding work elsewhere declines. Correspondingly, migration speeds up in
periods of growth (MacKay, 1992).Years of slow economic growth, and major
job losses in the economy as a whole, are crucial factors limiting movement out of weak or declining local economies, as the case of Newcastle
will demonstrate.
A l t h o u g h n e o - c l a s s i c a l a p p r o a c h e s focus on e c o n o m i c factors,
behavioural and motivational models have included other factors, such as
the importance of social networks in migration. The role of social networks
in e m e r g i n g u r b a n - r u r a l divides in South Africa is e m p h a s i s e d by
Wittenburg (1999), and by Cross (1999), who describe regional 'cells' of
migration, limiting free access to cities. While critiquing the myopia of neoclassical equilibrium models in focusing purely on economic factors,
Morrison (1993) models the effects of violence, using the behavioural concept of utility maximisation in violence-torn Guatemala. Although he suggests that violence is of secondary importance to economic factors, 'a narrow
SO UTH AFRICAN URBANISATION DYNAMICS
11
focus on economic factors is inappropriate in high violence countries' (p.
828). The significance of violence in shaping migration is recognised by a
number of authors in the South African context.
Structuralist and Post-Structuralist Perspectives
Major critiques of neo-classical approaches to migration have emerged
from both structuralist perspectives and from post-structuralist theorists,
and these also provide important insights into why movement out of economically weak or declining areas might be more limited than expected.
The important contribution of structuralist writers is to examine the way in
which macro-forces shape the structure of opportunities and constraints,
in terms of which migration decisions are made. Authors such as Auty (1995)
and Roberts (1989) have linked the nature of economic development to
patterns and forms of migration and urbanisation.
The modernisation assumptions contained in the urban transition model,
which underpins much neo-classical migration theory, and the expectation
of a unilinear development path, have been widely critiqued (Mabin, 1990).
While urban transition models might be appropriate in conditions where
the kind of e c o n o m i c d e v e l o p m e n t is similar to that anticipated by
modernisation theory, it is clearly not appropriate to many developing countries experiencing economic crisis and restructuring. Roberts (1989) for example, points to the emergence of fragmented migration patterns tied to
trends towards informalisation and casualisation in Latin America, which
in turn reflect processes of economic restructuring.
Migration patterns are therefore likely to vary according to the (changing) trajectory of economic development in particular countries. They cannot simply be derived from ideal models. Within particular countries,
changing patterns of uneven development will result in complex and often
unstable patterns of economic development over space. Nor is a neat mapping of patterns of economic development to migration, appropriate. Gould
(1995) shows the variations in the changing form of migration in East African countries since structural adjustment. While the different trajectories
of economic development in these countries account for part of these variations, Gould (1995) comments that there is little evidence for economic
determinism in migration systems, although economic motivations for migration may be strong. Similarly, Ferguson (1990) shows the complex mix
of forms of mobility within various periods in the Zambian Copperbelt. In
the 1980s, retrenched mineworkers returned to rural areas in the context of
difficult economic circumstances in urban areas. Even workers who might
have been seen as permanently urbanised in the 1960s and 1970s, and who
had limited rural connections moved there--essentially as a response to survival needs.
12
URBAN FORUM
At a lower level of abstraction, structuralists have examined the way in
which labour markets--and particularly labour market segmentation--provide differential opportunities and access for potential migrants, thus influencing patterns of migration (Flowerdew, 1992; Gordon, 1995). As
Johnson and Salt (1990) point out, migration is linked to the supply and demand for particular occupations, rather than to the situation of the labour force
as a whole. Social networks also play powerful roles in enabling access to
certain kinds of employment, and in organising migration in some contexts (Skeldon, 1990). Studies in Britain and the USA by Kitching (1992),
Owen and Green (1992), and Hodge (1985), suggest that the unskilled and
the poor are often disadvantaged in searching for work outside of their
home labour market, since new jobs are often sourced through word of
mouth, so contacts and networks within a local area are important. The
unemployed are also more familiar with opportunities for casual employment in their home town.
The forces underlying what is termed 'rural push' in neo-classical theories has also been a significant focus for attention by structuralist writers.
Rather than a simple push, rural out-migration is viewed as a complex and
often violent process, not necessarily in equilibrium with 'urban pull'. Chant
(1991a) for ex~imple, shows how households in Costa Rica have moved off
farms to economically stagnant small towns, where there are few opportunities available. While rural 'push' due to declining economic opportunities in agriculture has led to movement off the land, reasons for movement
to small towns relate to reproduction rather than employment possibilities. Households use these places as reproductive bases, from which males
engage in circular migration to larger cities. The availability of services and
infrastructure provides a more important source of attraction than does
narrowly economic opportunities.
Similarly in South Africa, as noted above, there is evidence in the 1980s
and 1990s of displaced farm workers moving (as a result of force or to
escape oppressive conditions on farms) to economically declining or
stagnant towns. In some cases, research indicates that forms of circular
migration continue from the town (Jones, 1994; Marcus et al, 1994).
Although these patterns might be described in the neo-classical literature
as a form of stage migration, in terms of which h o u s e h o l d s move to
towns close to them, the towns themselves largely serve as a site for reproduction. This is far from the narrow economic logic of the equilibrium
approach.
Beyond structuralist accounts, Boyle and Halfacree (1993) argue that
migration has to be seen within the context of individual biographies, rather
than as an action in time. Practical consciousness, perceptions, long term
intentions and multiple reasons for migration should be examined, rather
than single determinants. Further, migration occurs within a culturally con-
SO UTH AFRICAN URBANISATION DYNAMICS
13
structed context and is imbued with particular meanings, which may vary
for particular groups of people (Moon, 1995).
Households and Livelihood Strategies
The literature on households and livelihood strategies also provides an
i m p o r t a n t lens for e x a m i n i n g migration, and for u n d e r s t a n d i n g the
intersection b e t w e e n structural forces and micro-level responses. This
literature suggests that for low-income households, livelihood strategies
are likely to be multiple and diverse, and they are likely to change. Both
production and reproduction are significant parts of livelihood strategies,
and both mobility and immobility can be used to further survival. Migration
assists the poor to engage in a diversity of economic activities over space
(Schmink, 1984). Wood (1981) sees migration as a response to the failure of
other initiatives to achieve required income levels, but this formulation has
been critiqued as too narrow (Bach and Schramm, 1982). Rather, migration
can be used in more varied ways within multiple livelihood strategies and
can be aimed at either survival, or accumulation (Roysten, 1991). Migration
is also conditional on the division of labour within households and the
ability therefore of households to release members for this purpose (Chant
and Radcliffe, 1992). This in turn is linked to the organisation of production
and r e p r o d u c t i o n in the h o u s e h o l d and to p o w e r r e l a t i o n s w i t h i n
households (Chant and Radcliffe, 1992). The prospects for migration are
therefore influenced by social relations within households, by the demands
of reproduction, as well as by the external conditions that are more usually
emphasised.
Moser's (1996) comparative work on the importance of 'assets' in the
way households respond to economic crisis provides further insights into
the dynamics of mobility and immobility. Firstly, a common response to
economic decline is to increase the number of household members who
work in the formal and informal sectors. Increasing women's employment
is most common, but in some cases, child labour is extended as well. The
increase in numbers of those employed within households, however, may
be insufficient to make up for decline in income from the loss of better
paid (often men's) work, and may be coupled with consumption and budget cuts 7. Secondly, households diversify their income sources through housi n g - t h r o u g h h o m e - b a s e d enterprises, rental, and by using housing to
extend reciprocal relations. Thirdly, a common response to crisis is to increase reliance on extended family support networks, and to reorganise the
household. Fourth, increasing labour migration (including international
labour migration), often on a partial or circular basis, coupled with remittances, can be an important response to crisis. Social networks are likely to
be drawn on in this process. Fifth, economic pressure may lead to increas-
14
URBAN FORUM
ing reliance on informal networks and relations of reciprocity, although
crisis may also erode these 'stocks of social capital', reducing the extent to
which reciprocal relations can be mobilised to assist households.
Although Moser (1996) and others do not spell out how these processes
might link to the role of place in livelihood strategies, there are clear pointers. First, the role of social networks and reciprocity is emphasised by a
number of authors. Although these relations may occur over space, they
also occur within places. Second, ties to place may be consolidated through
security and investment in housing. To the extent that housing can be seen
as a productive asset--and also an important form of security--it will not
easily be abandoned. Housing represents a considerable investment which
might have occurred w h e n households had resources (such as in better
times, or in more favourable stages of the life cycle of the household), and
cannot easily be recreated in another place. Third, the possibility of extending the use of household labour may be a way in which households
avoid wholesale migration, and consolidate resources in places, even if overall incomes are lower. Local employment might be preferred to circular
migration since households are likely to retain a greater proportion of income earned. Restructuring may therefore not be accompanied by increasing u n e m p l o y m e n t in particular places or out-migration so much as by
declining net incomes. Fourth, partial or circular forms of migration may
be a way in which income is extended and risk reduced, but this involves
holding on to places where there is a level of security. Migration is nevertheless one possibility within livelihood strategies and one response to cris i s - o t h e r s discussed here may be used instead of or in conjunction with
it. Fifth, places play important roles in reproduction. To assume then, as do
the equilibrium models, that there is an easy and automatic shift from places
experiencing decline or restructuring to others where wages, incomes, and
employment levels are higher, overlooks the roles that places might play in
survival, and how past investments in social and physical capital may mediate the possibility and likelihood of migration as an option--particularly
in a context of national economic recession or crisis.
Gender and Migration
Finally, the literature on gender and migration provides further insights
into why people might not necessarily move in response to economic
stagnation, restructuring or decline. Gendered patterns of migration are
explained in terms of gender power relations, including intra-household
decision-making and resource allocation, and by socially constructed gender
segregation of labour markets. Women' s migration is often constrained by
their responsibility for child-care, and by socially constructed restrictions
on their movement (Chant and Radcliffe, 1992). Studies have for example
S 0 UTH AFRICAN URBANISATION DYNAMICS
15
documented the constraints which married women face (e. g. see Jones,
1994). While unmarried women may be freer, responsibility for child-care
can limit migration or lead to a preference for local employment, even if it
is inferior. But these patterns are not necessarily static. The feminisation of
the labour force (Standing, 1989), for example, has been associated with
rising female migration. Further the gender dynamics of mobility and
immobility can reflect gender power relations, which are not immutable. Thus
gender relations might provide a powerful reason why (some) women remain
trapped in places, but women's immobility cannot be assumed a priori.
Several theories thus provide insights into why people might move into,
or remain in, places experiencing economic stagnation, decline or restructuring. The following section explores the case of Newcastle, using some of
these theories, to show the limits of normalisation assumptions.
THE CASE OF NEWCASTLE
Newcastle is a town in North-West KwaZulu-Natal that grew enormously
as a consequence of apartheid policies of forced removals, influx control,
and industrial decentralisation. The population of the town grew from 17
554 in 1960, to 352 925 in 1991, with growth rates of 10 percent p. a. in the
1960s, 17 percent in the 1970s, and 3.4 percent in the 1980s. Although the
establishment of the third iron and steel works of the parastatal Iscor in
the town in the town in 1972 led to major economic growth in the town,
with employment in manufacturing increasing from 5796 to 20 102 from
1970 to 1980, while total employment grew from 31 860 to 57 711 over the
same period, population growth far outstripped employment growth.
Forced removals were particularly intense in North Western Natal, and
at least 50000 African people were removed to Madadeni and Osizweni,
which were created as regional relocation townships. Removals from the
town of Newcastle were also significant. Many African people also moved
to these places, and to the informal settlement of Blaaubosch, to escape
oppressive conditions on farms. Since they fell within a homeland, movement into the area was far easier than to Gauteng, although it was not
unrestricted. Agricultural restructuring and farm evictions also played a role
in fuelling population growth. A substantial part of the growth of the area,
therefore, was not linked to economic development of the town, although
people increasingly came to depend on employment generated there.
The establishment of Iscor in the town, and the subsequent growth of
the local economy, enabled households from rural areas, and those who
had relied on migrant labour, to stabilise in the area and to centre their
economic strategies on the town. Households have increasingly localised
their livelihood strategies. Nevertheless, by 1992, levels of migrancy were
16
URBAN FORUM
high for an urban area, with some 23 percent of the African labour force
involved in migrant work elsewhere, and with a range of 14 percent of
households in Madadeni, 30 percent in Osizweni, and 49 percent of households in Blaaubosch receiving some form of migrant income (DRA, 1992).
Although the location of Iscor led to boom conditions in the town, this
was short-lived. Several rounds of restructuring have occurred since 1982,
as Iscor responded to recession and an increasingly competitive international market. Employment in Iscor declined from 13 000 in 1977 to 8080 in
1989, to 3000 in 1999. While the first round of retrenchments in the early
1980s particularly affected migrants, it is largely locals who have borne the
brunt of downsizing since then. Other heavy industries also downsized,
particularly in the 1990s (Todes, 1999). From the mid-1980s, employment
in the town was shored up the growth of the clothing industry, as Taiwanese firms established in the area, and as South African firms moved in response to competitive pressures, escaping rising wages and growing
unionisation in the cities. These movements were facilitated by the generous
industrial decentralisation incentives on offer, and by local initiatives to bring
in Asian firms to the area. By 1991, manufacturing was one of the few sectors
where employment had not declined, but there was a shift to low wage employment, and a change in the gender composition of employment.
Employment declines accelerated in the 1990s, with 17 percnet of jobs
lost between 1990 and 1994 s (Todes, 1999). Decline was particularly concentrated in heavy industry, where 29 percent of jobs were lost, but 4 percnet
of jobs were also lost in the clothing industry, as growing tensions over
labour relations, and rising import penetration affected the industry. The
path of low waged industrialisation is increasingly fragile. Industrial decline seems to have affected other areas of the economy as well, although
the town has by no means collapsed (Todes, 1997).
Despite restructuring, the area experienced net in-migration in the 1980s.
People continued to move into the area in the 1990s, albeit at much lower
rates (Todes, 1997), and there is no evidence of significant out-migration 9
These trends are contrary to the expectations of the normalisation thesis,
although growth rates have tapered off, indicating responsiveness to change.
Research on migration dynamics in the area TM showed that the migration
into the area could be explained by continuing removals from farms, which
persisted into the 1990s, and by the effects of violence in Gauteng, which
led to a level of migration back to more peaceful smaller towns, particularly under conditions of political polarisation in the late 1980s and early
1990s. Economic restructuring also led to a level of in-migration, as workers moved in to take up employment in clothing, although these effects
were rather limited.
The research shed light on why people have remained in the area, despite restructuring. Crime and violence has played a significant role in shap-
S O U T H AFRICAN URBANISATION DYNAMICS
17
ing migration. While Newcastle did not escape political violence, it was
less pervasive there than many areas. Respondents interviewed saw it as a
safer place than the cities. Older retrenched Iscor workers were particularly wary of moving to Gauteng or other cities.
Housing plays a critical role of the consolidation of households in place.
The issue is not simply the limited housing market which exists in townships, but the fact that years of effort and considerable household resources
have been spent building houses. In the apartheid years, investment in
housing was possible in h o m e l a n d areas in contrast to other areas, and a
high level of consolidation occurred in townships and informal settlements
around Newcastle. Households which were tied into spatially extended livelihood strategies had received remittances specifically for housing. Osizweni,
which began as a self-help township in the 1960s, now has very substantial
houses. Similarly, in Blaaubosch, the devastating effects of the floods in 1984
led to an agreement that households could build permanent structures on land
owned by landlords. Under apartheid, the supply of sites and housing in the
area was more plentiful than in many other areas of KwaZulu-Natal, and
housing conditions are generally better than in metropolitan areas.
Immobility of households in Newcastle is also the result of limited access to employment within cities. It is well k n o w n that the labour absorption capacity of the South African economy has been declining, and that an
increasing proportion of people are occupied in the informal sector. In the
absence of secure well-paid work in cities, households are retaining the
securities they have in place. In the case of Newcastle, there were numerous instances of household members who had tried to migrate to cities,
but finding no work, had returned. Only 21 percent of the 104 unemployed,
retrenched heavy industry workers surveyed had not attempted to look for
work elsewhere. Some 56 percent had sought work elsewhere, but not finding it, returned. The remainder had only been able to find temporary or
part-time work, or had returned after losing a job, or for personal reasons.
Only 7 percent were still looking for jobs in the cities.
Mobility is also proscribed by the costs of seeking a job (transport and
accommodation), and by access to contacts and accommodation in metropolitan areas. As standard migration theory would suggest, households
which are too poor (but still have some security) found it difficult to leave.
Within households, the costs of temporary migration meant that it was used
selectively as a survival strategy. While some members of the household
might have gone to metropolitan areas in search of jobs, others who were
unemployed remained behind due to the costs of job search.
The research demonstrated the constraints on w o m e n ' s movement. It
was rare for w o m e n to experience the freedom of movement that men did.
Their movement was much more contingent and d e p e n d e n t on access to
contacts and networks. Women's mobility varied according to their posi-
18
URBAN FORUM
tion in the household. Married w o m e n could not move at w i l l - - t h e i r
husband's power in this regard was clearly apparent. Unmarried women
were freer to move, but this depended on their position and conditions
within the household. They were frequently constrained by their roles as
care-givers--responsibility for children, the sick and disabled, and for old
parents.Young unmarried women, especially those without children, could
move more easily, although the support expected from daughters sometimes meant that migration was not as easy as for sons.
Despite high levels of unemployment among retrenched heavy industry
workers, their households have largely remained in Newcastle, reliant increasingly on women's work, and diverse sources of local and migrant income. Some 55 p e r c e n t of r e t r e n c h e d heavy industry workers were
unemployed, and a further 9 percent were retired. Only 18% had found a
job locally, and another 14 percent were self-employed. Nevertheless, a
minority of retrenched worker households had no evident sources of income (6%) or relied entirely on pensions (8%). Households have survived
by extending the numbers of those employed, by multiple livelihood strategies, involving a combination of incomes, pensions, commuter income and
remittances. Circular migration played a part in this, with 8 percent of workers becoming migrants, and some 27 percent of households having income
from remittances or commuting, but this was usually one source amongst
others. Women's participation in the labour force was a significant way in
which retrenched worker households survived, with 80 percent of households having women contribute to household income, compared to 55 percent where men did. The clothing industry however provided a limited part
of this employment, with only 22 percent of retrenched worker households
containing clothing workers.
The Newcastle study did not find a complete disjuncture b e t w e e n
urbanisation and economic d e v e l o p m e n t , and certainly conditions in
Newcastle are far better than they are in rural areas, in isolated resettlement areas, and in small declining industrial decentralisation points. It has
nevertheless shown the limits of assumptions that migration is necessarily
a corrective to local economic stagnation or decline, and to the pattern of
settlement created under apartheid--at least if current trends in employment creation continue.
CONCLUSION
The study has shown that while there is evidence that some people are
moving to the cities and other centres of growth, there are also important
exceptions, with people staying in or even moving to places with weak or
SO U T H A F R I C A N U R B A N I S A T I O N D Y N A M I C S
19
declining economic bases. Past policy played a role in underpinning these
s e t t l e m e n t p a t t e r n s - - p a r t i c u l a r l y by s h a p i n g the m a t e r i a l basis of
urbanisation through subsidies, the development of sites and housing, and
by influencing the spatial structure of economic opportunities--but it is
not clear that policy shifts in these spheres will automatically translate into
movement.
A key issue is the current form of economic growth, which is characterised
by declining formal employment and a shift towards the use of more skilled
workers, coupled with the proliferation of marginal survivalist informal
businesses. Under these circumstances, those who have some assets and
security in place are likely to hold on to them, relying on multiple livelihood strategies, including pensions, local sources of income, coupled with
forms of commuting or circular migration, and migration by individuals. In
contrast to the traditional dualism of people versus place, place may play
important roles in survival amongst low-income households. Those who
are desperate or destitute are likely to move, but some may be too poor, or
may not have the social networks, to enable them to do so. High rates of
employment creation, conversely, could result in rapid migration to centres
of job growth--but even under these circumstances, movement of entire
households from places with weak or declining economic bases is likely to
take a generation. Current trends, however, are not promising, with further
employment declines being reported.
Movement to cities is also limited by perceptions of crime and violence
in cities, and differential housing conditions. While conditions could change
in future, the extent of crime and housing backlogs, and the complexity of
dealing with these issues in metropolitan areas, suggests that they are unlikely to be resolved quickly.
These findings raise questions about policy positions that assume that
normalisation can and will easily occur, and the 'workers to work' logic.
There is much to support policies that facilitate growth, and enable
urbanisation in metropolitan areas, and other centres of economic growth.
Certainly, these places will have to accommodate significant urbanisation
in future, and many such places have substantial backlogs in housing, infrastructure and services, as a consequence of the anti-metropolitan policies of the past. Ironically, these have been perpetuated in some provinces,
such as the Western Cape, post apartheid. Nevertheless, the complex and
messy legacy of apartheid settlement patterns is likely to remain for some
time, and cannot simply be wished away. In this context, a sole focus on
metropolitan areas, and other places demonstrating comparative or competitive advantages is problematic. For instance, large populous centres with
significant--but declining--economic bases, such as Newcastle, may not
be the powerhouses of the future, but they contain resources and strengths
that can be built on. Their sustainability is important to people who have
20
URBAN FORUM
assets such as housing and social networks there, and they offer greater
opportunity for survival and for the development of livelihoods than rural
areas.
This does not imply a return to industrial decentralisation policies, or an
advocacy to attempt to create unsustainable forms of development in unlikely places. Rather, there is a need to move beyond the somewhat polarised
positions that have often characterised policy debates - emphasising either
urban or rural, core or periphery, efficiency or equity, people or place, to
contextually appropriate interventions, based on an understanding of local
conditions, needs and potentials. For instance, investment in housing and
related infrastructure may be more important in areas of growth than in
places where the local economy is stagnant, and where initiatives to secure
parts of local industry may be more significant. Similarly, given the extent
and significance of rural poverty, rural development would seem to be crucial, but should not be counterposed against urban development, or be seen
as a way of preventing movement to urban areas. Further, many resettlement areas created in the past may need to be accepted as permanent.Yet
there are likely to be significant variations between these settlements, in
terms of their social and economic dynamics, and the resources, investments and networks people have there. Policy needs to be sensitive to local
social contexts, to avoid blindly reinforcing--or destroying--such places
through administrative actions. Similarly, while disjunctures created in the
past, between where people live, and where jobs are, may have to be accepted, there is little reason to create them anew. There is thus a need to
move beyond the rather totalising narrative of the normalisation thesis,
and narrow conceptions of policy.
NOTES
1. This section draws on research conducted for the Spatial Guidelines for
Infrastructure Investment and Development project, Office of the Presidency. Funding for this work is acknowledged. The views and arguments
expressed here, however, are my own, and cannot be attributed to that
Office or to the SGIIDS project.
2. There are severe problems in comparing the 1996 census with previous
censuses, as a consequence of differences in the way they have been
corrected.
3. Cross et al (1996a, b, 1998) Kiepel (1994); Hindson and McCarthy (1994);
Todes (1994, 1997), Meth (1998)
4. Cross et al (1999); Bank (1997), Bekker (1999), Spiegel et al (1996) Dewar
et al (1990)
SO UTH AFRICAN URBANISATION DYNAMICS
21
5. Krige (1995, 1996, 1998), Tomlinson and Krige (1997)
6. Sites for the location of African people forcibly removed under Apartheid. It is estimated that some 3,5m people were removed (including
coloured and Indian people removed under the Group Areas Act). The
policy displaced African owners and tenants from both urban and rural
areas designated for whites, and was also used to reduce the number of
African workers living on farms owned by whites. While some areas
that people were removed to were within towns and cities, others were
in far flung areas, or required lengthy commuting to the nearest town.
The term 'resettlement areas' is used to refer to these more distant
areas.
7. Chant' s (1991b) work on Mexico corroborates these points.
8. Figures are based on a survey of close to 100 percent of industries conducted together with Gillian Hart, University of Berkeley.
9. By 1991, e m p l o y m e n t rates were much lower than in Durban and
Gauteng, and unemployment rates were higher. On DBSA (1995) figures, labour force participation rates in Newcastle were 55.5 percent,
compared to 64. 9 percent in Durban, and 74. 9 percent in Gauteng.
Unemployment rates were 25 percent in Newcastle, compared to 16
percent in Gauteng and 21 percent in Durban. Conditions are however
far better than in rural areas.
10. Both qualitative and quantitative methods were used. In-depth qualitative interviews were conducted with 31 households in 1994/5, while a
larger survey of 354 households was conducted in 1995/6. This study
included 101 households containing retrenched workers. These studies
are more fully reported in Todes (1997) and Todes (1998a and b)
REFERENCES
Auty, R., 1995: Patterns of Development: Resources, Policy and Economic Growth,
Edward Arnold, London.
Bach, R. and Schram, L., 1982: Migration crisis and theoretical conflict, International Migration Review, 16 (2), 320-41.
Bank, L. 1997: 'Moving Out of Step'. Urbanisation, Migration and Livelihoods in the Eastern Cape, Institute for Social and Economic Research
Report, Rhodes University, Grahamstown.
Bekker, S. 1999: A Profile of Urbanisation in the Western Cape, Unpublished Draft paper.
CDE (Centre for Development and Enterprise), 1995: Post-Apartheid Demographic and Income Distribution Model, CDE, Johannnesburg.
CDE (Centre for Development and Enterprise), 1996: South Africa's Small
Towns: New Strategies for Growth and Development, CDE, Johannesburg.
22
URBAN FORUM
CDE (Centre for Development and Enterprise), 1998: South Africa's 'Discarded People':
Survival, Adaptation and Current Changes, CDE, Johannesburg.
Chant, S., 1991a: Gender, migration and urban development in Costa Rica:
the case of Guanacaste, Geoforum, 22, 237-253.
Chant, S., 1991b: Women and Survival in Mexican Cities. Perspectives on Gender, Labour Markets and Low-Income Households, Manchester University
Press, Manchester.
Chant, S. and Radcliffe, S, 1992: Migration and development: the importance of gender, in Chant, S. (ed.), Gender and Migration in Developing
Countries, Belhaven, London, 1-29.
Charney, A., 1993: Migration and the public sector: a survey, Regional Studies, 27, 313-326.
Cross, C. 1999: Spatial implications of population migration for South
Africa's development and infrastructure programmes, Unpublished Report for the Spatial Guidelines for Infrastructure, Investment and Development, for the Coordination and Information Unit, Office of the
Deputy President.
Cross, C., Luckin, L., Mzimela, T., and Clark, C., 1996: On the edge: poverty, livelihoods and natural resources in rural KwaZulu-Natal, in Lipton,
M., Ellis, F., and Lipton, M. (eds), Land, Labour and Livelihoods in Rural
South Africa, Indicator Press, Durban, 173-214.
Cross, C., Mngadi, T., Mbhele, T., Mlambo, N., Kleinbooi, K., Saayman, L.,
Pretorius, H. and Bekker,S., 1998: An Unstable Balance: Migration, Small Farming, Infrastructure and Livelihoods on the Eastern Seaboard: KwaZulu-Natal,
Unpublished Report for the Development Bank of Southern Africa.
Cross, C., Bekker, S., Mlambo, N., Kleinbooi, K., Saayman, Pretorius, H.
Mngadi, T., Mbhele, T., 1999: An Unstable Balance: Migration, Small
Farming, Infrastructure and Livelihoods on the Eastern Seaboard: Eastern Cape and Western Cape, Unpublished Report for the Development
Bank of Southern Africa.
Data Research Africa (DRA), 1992: Income and Expenditure In KwaZuluNatal, Unpublished Report for the KwaZulu Finance and Investment
Corporation, Durban.
Davies, R. and Head, J. 1995:The future of mine migrancy in the context of broader
trends in migration in South Africa,Journal of Southern African Studies, 21 (3).
Development Bank of Southern Africa (DBSA), 1995: KwaZulu-Natal: Statistical Macro-Economic Overview, Development Information Paper 100,
DBSA, Midrand.
Dewar, D. 1996: Small Towns In Development: A South African Perspective,
Development Paper 122, DBSA, Midrand..
Dewar, D., Rosmarin, T., and Watson, V., 1991: Movement Patterns of the African Population in Cape Town: Some Policy Implications, Working Paper
No. 44, Urban Problems Research Unit, University of Cape Town.
SO UTH AFRICAN URBANISATION DYNAMICS
23
Ferguson, J., 1990: Mobile workers, modernist narratives: a critique of the
historiography of transition on the Zambian Copperbelt, Journal of Southern African Studies, 16 (3), 387-412, and 16 (4), 604-621.
Flowerdew, R., 1992: Labour market operation and geographical mobility,
in Champion, T. and Fielding, T. (eds), Migration Processes and Prospects.
Volume 1. Research Progress and Prospects, Belhaven, London, 135-147.
Gordon, I., 1995: Migration in a segmented labour market, Transactions of
the Institute of British Geographers, 20, 139-155.
Gould, W., 1995: Migration and recent economic change in East Africa, in
Baker, J., and Aina, T. (eds), The Migration Experience in Africa, Nordiska
Afrikainstituten, Uppsala, 122-147.
Halfacree, K. and Boyle, R, 1993: The challenge facing migration research:
the case for a biographical approach, Progress in Human Geography, 17, 333348.
Hart, G, and Todes, A., 1997: Industrial Decentralisation Revisited, Transformation, 32, 31-53.
Hindson, D. and McCarthy, J., 1994: Defining and gauging the problem, in
Hindson, D. and McCarthy, J. (eds), Informal Settlements in KwaZuluNatal, Indicator Press, University of Natal, Durban, 1-28.
Hodge, I., 1985: Employment expectations and the costs of migration, Journal of Rural Studies, 1 (1), 45-57.
Johnson, J. and Salt, J., 1990: Labour migration: the general context, in
Johnson, J. and Salt, J. (eds), Labour Migration. The Internal Geographical
Mobility of Labour in the Developed World, David Fulton, London, 1-13.
Jones, B., 1994: Eastern Cape Province. Aberdeen Case Study, Unpublished
Report to the Land Reform Research Project.
Kiepel, J., 1994: A typology of informal settlements in KwaZulu-Natal, in
Hindson, D. and McCarthy, J. (eds), Here to Stay. Informal Settlements in
KwaZulu-Natal, Indicator Press, Durban, 123-132.
Kitching, R., 1990: Migration behaviour among the unemployed and lowskilled, in Johnson, J. and Salt, J. (eds), Labour Migration. The Internal
Geographical Mobility of Labour in the Developed World, David Fulton,
London, 172-190.
Krige, S. 1995: Demographic Profile of the Free State, Occasional Paper No.
15, Urban and Regional Planning, University of Orange Free State,
Bloemfontein.
Krige, S. 1996: Botshabelo: Former Fastest-Growing Urban Area in South Africa Approaching Zero Population Growth, Occasional Paper No. 20, Urban and Regional Planning, University of Orange Free State,
Bloemfontein.
Krige, S. 1998: The challenge of dismantling spatial patterns constructed
by apartheid in the Bloemfontein-Botshabelo-Thaba Nchu region, Acta
Academica Supplemtum 1.
24
URBAN FOR UM
Liebbrandt, M. and Woolard, I., 1996: Rural Labour Markets, Migrant Labour
and Rural Poverty in South Africa, Unpublished Report to the International Labour Organisation, Geneva.
Mabin, A., 1990: Limits of urban transition models in understanding South
African urbanization, Development Southern Africa, 7, 311-322.
Mabin, A. 1997: Imaginative responses to human settlements, Unpublished
Report for the White Paper on Local Government, Pretoria.
Mackay, R., 1992: Local labour markets, regional development and human
capital, Regional Studies, 27, 783-795.
Marcus, T., Eales, K., and Wildschut, A., 1996: Down to Earth. Land Demand
in the New South Africa, Indicator Press, Durban.
Meth, P. 1998: Rethinking the Dumping Grounds: The Case of Bilanyoni,
Unpublished PhD.Thesis, University of Cambridge.
Meth, P. 2000: Representation and formalisation: A settlement of contradictions, Urban Forum, 11, 31-48.
Milne, W., 1993: Macroeconomic influences on migration, Regional Studies,
27, 365-373.
Mokate, R. Schawbe, C., Makinta, V. and O'Donovan, M. 1999: Government spending theme paper: The spatial allocation of infrastructure
and development spending in South Africa, Unpublished Report for
the Spatial Guidelines for Infrastructure, Investment and Development,
for the Coordination and Information Unit, Office of the Deputy President.
Moon, B., 1995: Paradigms in migration research: exploring'moorings' as a
schema, Progress in Human Geography, 19, 504 -524.
Morris, M., Barnes, J., and Dunne, N. 1998: Globalisation and the restructuring of Durban's Industry, Unpublished Paper presented at the City
of Durban Conference, Durban.
Morrison, A., 1993: Violence or economics: what drives internal migration
in Guatemala, Economic Development and Cultural Change, 41, 817-831.
Moser, C., 1996: Confronting Crisis. A Comparative Study of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities, Environmentally Sustainable Development Studies and Monographs Series
No. 8, World Bank, Washington DC.
Owen, D. and Green, A., 1992: Migration patterns and trends, in Champion, T., and Fielding, T. (eds), Migration Processes and Prospects. Volume
1. Research Progress and Prospects, Belhaven, London, 17-38.
Roberts, B., 1989: Urbanisation, migration, and development, Sociological
Forum, 4, 665-691.
Roux, A., 1991: Migration and regional policy, Urban Forum, 2 (1), 41-58
Roysten, L., 1991: Persistent Circular Migration: Evidence, Explanations and
Planning Implications, Unpublished MSc Dissertation, University of the
Witwatersrand, Johannesburg.
SOUTH AFRICAN URBANISATION DYNAMICS
25
Schmink, M., 1984: Household economic strategies: review and research
agenda, Latin American Research Review, 19 (3), 87-102.
Skeldon, R., 1990: Population Mobility in Developing Countries: A Reinterpretation, Belhaven, London.
Spiegel, A., Watson, V., and Wilkinson, P. 1996: Devaluing diversity?: national housing policy and African household dynamics in Cape Town,
Urban Forum, 7 (1), 1-30.
Standing, G., 1989: Global feminisation through flexible labour, World Development, 17, 1077-1095.
Todaro, M., 1981: Economic Development in the Third World, Second Edition,
Longman, New York.
Todes, A. 1994: Urbanisation and urban management in KwaZulu-Natal,
Development Southern Africa, 11, 541-554.
Todes, A., 1997: Restructuring, Migration and Regional Policy in South Africa: The Case of Newcastle, Unpublished Ph.D. Thesis, University of
Natal, Durban.
Todes, A. 1998a: Gender, place, migration and regional policy in South Africa, in Mapetla, M., Larsson, A., and Schlyter, A., (eds.) Changing Gender Relations in Southern Africa. Issues of Urban Life, Institute for
Southern African Studies, National University of Lesotho, Roma.
Todes, A. 1998b:Socio-spatial effects of restructuring: the case of Newcastle,
Society in Transition, 29 (1-2), 40-57.
Todes, A. 1999: Industrial restructuring in South Africa. The case of
Newcastle, Tidjschrift voor Economische en Sociale Geografie, 90, 379-390.
Tomlinson, R., 1990: Urbanisation in Post-Apartheid South Africa, Unwin
Hyman, London.
Tomlinson, R. 1997: Gauteng Urbanisation Status Quo Report, Unpublished
Report for the Gauteng Provincial Government, Johannesburg..
Tomlinson, R. and Krige, S. 1997: Botshabelo: coping with the consequences of
urban apartheid, InternationalJournal of Urban and RegionalResearch,21, 691-705.
Urban Foundation, 1989: Population Trends, Policies for a New Urban Future Series No.l, Urban Foundation, Johannesburg.
Urban Foundation, 1990: Regional Development Reconsidered, Policies for a
New Urban Future Series No.2, Urban Foundation, Johannesburg.
Urban Foundation, 1994: Outside the Metropolis: The Future of South Africa's
Secondary Cities, Urban Foundation, Johannesburg.
Urban Strategy Department, 1995: Settlement Areas and Population Estimate Project, Durban Metropolitan Area, 1995, Unpublished Report,
Corporate Services, City of Durban.
Wittenburg, M. 1999: A spatial analysis of the labour market, Unpublished
Report for the Spatial Guidelines for Infrastructure, Investment and
Development, for the Coordination and Information Unit, Office of the
Deputy President.
26
URBAN FORUM
Wolpe, R. 1997: Reality Check: An Overview of Non-Metropolitan Local
government in South Africa, Unpublished Report to the Local Government White Paper Committee, Cape Town..
Wolpert, J., 1965: Behavioural aspects of the decision to migrate, Papers of
the Regional Science Association, 15, 159 -72.
Wood, C., 1981: Structural change and household strategies: a conceptual
framework for the study of rural migration, Human Organisation, 40 ,
338-44.