skip to Main Content

TRANSPORT PROBLEMS ASSOCIATED WITH POVERTY IN SOUTH AFRICA

TRANSPORT PROBLEMS ASSOCIATED WITH POVERTY IN SOUTH AFRICA

L A KANE

Urban Transport Research Group

Department of Civil Engineering, University of Cape Town,

Private Bag, Rondebosch, 7701

ABSTRACT

Since 1994 the eradication of poverty and the redressing of inequalities in South African
society has been a central theme of the South African government. Despite its high profile,
an international standard definition of poverty is not available. For this paper ‘poverty’ is seen
as a relative concept, and the view is taken that we should not be considering one South
African poverty problem, but rather developing strategies for the metropolitan, urban and the
rural poor. Accordingly, the analysis in the paper subdivides South African the population into
metropolitan, urban and rural households, and then further subdivides households into
‘quintiles’ according to income. Detailed analysis of a broad range of transport issues follows:
overall amount of travel; travel purpose; public transport accessibility, affordability and
attitudes of customers; vehicle ownership; accessibility to shopping, education and medical
facilities.
The first major theme to emerge from the analysis is that of difference, between the rural and
metro/urban poverty experiences. Secondly, there is a theme of better accessibility for the
urban poor than for the metropolitan poor. The third theme is the overarching importance to
the poor of the walking trip. The fourth theme is an overall pattern of decreasing accessibility
with decreasing income. This was particularly true in metropolitan areas. Finally the
important role of the minibus taxi to the poor was clear. As a vehicular mode it dominated in
the poor sector.

In conclusion the paper argues for transport planning policies directed to the poor in the so-
called ‘Second Economy’. At a practical level transport planning for this economy would:

 Provide safe, secure, direct, well maintained walk and walk/cycle paths, and road-
crossing facilities, especially to schools.

 Improve bus routing between informal areas and key services
 Partner with education authorities with respect to schools locations, not only in rural
areas but also in metros
 Partner with police and national government with respect to the regulation and
enforcement of improved safety standards for taxi vehicles and taxi driving
 Invest in the upgrading of taxi terminals

3
1. POVERTY IN A SOUTH AFRICAN CONTEXT
1.1 Introduction
Since 1994 the eradication of poverty and the redressing of inequalities in South African
society has been a central theme of the South African government. In 2005, as part of his
presidential address at the state opening of parliament President Mbeki stated:
“we must achieve new and decisive advances towards [amongst
others]…eradicating poverty and underdevelopment, within the context of a
thriving and growing First Economy and the successful transformation of the
Second Economy…These objectives constitute the central architecture of our
policies and programmes, intended to ensure that South Africa truly belongs
to all who live in it, black and white.”
In this context of a policy drive towards poverty eradication, the nature of poverty needs to be
discussed. Any international definition of it remains lacking, despite recent high profile events
such as the ‘Make Poverty History’ campaign in the UK, and the work of the United Nations
in establishing the Millennium Development Goals. The only agreement in academic circles
regarding the nature of poverty, is that “poverty is a contested concept” (Noble et al, 2004).
Noble argues that it is contested for good reason. Ultimately, he argues, the definition of
poverty is political, and it reflects the values of society, by establishing a norm for what is
acceptable, or not. One implication of this argument is that definitions of poverty are tied in a
country’s particular circumstances and culture. The ‘poor’ of Europe are not the same as the
‘poor’ of Africa, and hence poverty definitions need to be made locally, and in relation to the
current norms of that society.
1.2 Towards a working definition of poverty
Parnell and Mosdell (2003) and Noble, Ratcliffe and Wright (2004) have worked in the field of
poverty in South Africa, and have contrasted the various approaches to defining poverty.
They suggest that there are several approaches, ranging from a predefined measure of
income, to a consideration of basic needs, to ideas about human rights and access to a
healthy environment. The approaches vary from the simplistic to the complex, with varying
degrees of consideration of economic, environmental, social and human dimensions. The
analysis of the 2003 first National Household Travel Survey (NHTS) (DoT, 2005) which
follows takes the following position with regard to defining poverty:
Poverty is a relative concept, that is, the poor cannot be easily defined using
an absolute income measure, as this will vary widely across rural and urban
contexts. Urban living is more expensive than rural living for various reasons
(see Parnell, 2003) and we should not be considering one South African
poverty problem, but rather developing strategies for the urban and the rural
poor.
Accordingly, the analysis which follows subdivides the South African population into
metropolitan, urban and rural households, and then further subdivides households into
approximately 20% sized subsets of data (‘quintiles’) according to income.

4

1.3 Analysis method
Full details of how the ‘quintiles’ were developed, their exact sizes and income ranges are
discussed fully in the appendix. The main points to note are:
 The metropolitan, urban and rural populations of South Africa are not the same size.
The metro areas comprise 4.56 million households; the urban areas comprise 3.48
million households and the rural areas comprise 4.39 million households;
 although described as ‘quintiles’ the data subsets are not exactly 20% of each area’s
households. Some so-called ‘quintiles’ are larger, or smaller than one fifth of the total
survey population minus missing data. Full details why are given in the appendix to
the paper.
 The figure below illustrates the ‘quintile’ size discrepancy by area. Although quite
significant in places, it is the best possible solution given the analysis method selected
for this paper, and still provides a well sized subset of poor households for detailed
analysis.

Number of households per ‘quintile’ used in

analysis

0 1 2 3 4 5
Metro
Urban
Rural
Area

Households (millions)

Poor Poor-middle Middle Middle-rich Rich Missing

2. NHTS MAIN OUTCOMES WITH RESPECT TO POVERTY
2.1 Introduction
As transport planners we tend to talk in a fairly standard way about transport. Trips, trip
purposes and mode splits are all familiar terms in the planners’ vocabulary. In the next
section of the paper these standard survey outcomes are described for South Africa, with
particular emphasis on the poor. Following this section are a series of sections which
consider accessibility to, and use of, public transport, and then the accessibility of selected
trip purposes. In each section the emphasis is on the poorest 20%, the poor income ‘quintile’
of each metropolitan, urban or rural population.

5

2.2 Amount of travel
Although numbers of trips and trip
lengths were not directly measured by
the NHTS, an estimate of the overall
amount of travel is possible. The figure
opposite shows the percentage of
people who traveled on the survey day,
by all modes including walking, by
income quintiles. Two interesting points
emerge: firstly the fewer trips taken
overall in the rural setting. This may be
attributed to difficulty in access: longer
walk times to some destinations, and
lack of affordable or reasonable
motorized transport for others in rural
areas, as well as fewer opportunities for

Percent traveling on survey day by income and

area

60.0%
65.0%
70.0%
75.0%
80.0%
85.0%
90.0%
95.0%

Poor Middle Rich

Income quintile Percent

Metro Urban Rural

work and shopping.
Secondly, the general trend is for an increase in metro/urban trip-making activity as incomes
increase, with the rich traveling 20% more than the poor in metropolitan areas, and 10% more
in urban areas.
2.3 Purpose of travel
Percent travelling by purpose on survey day – All and poor metropolitan

0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%

Shopping Visiting Education Work
Purpose Percent
Poor All

Percent travelling by purpose on survey day – Poor: metropolitan and rural

Shopping
Visiting Visiting
Education

Work

Work

Shopping

Education

0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%

Metro Rural
Area Percent

For the poor in metropolitan areas the most popular purposes were shopping (28%); visiting
(25%); and education (23%). Urban areas show a similar pattern. For all income groups
combined, however, the most popular purposes are distinctly different with work and
shopping being the most popular purposes in metropolitan areas (29%), followed by
education (26%) and visiting (23%). Again urban areas follow similarly.
Notably in the rural areas, although there are fewer trips overall, the most popular trip

6

purpose is education, with 38% of the poor reporting education trips; followed by visiting
(23%) and shopping (19%).
Visiting is the second most important
purpose for the poor, in all areas, and
accounted for approximately one quarter of
all trips. According writers on poverty, social
networks are an important asset of the poor,
and one which protects them against
hardship, through, for example: sharing of
chores; informal care for sick, elderly and
children; and informal lending of food and
money. The graph opposite shows the
decreasing importance of the visiting trip as
incomes increase.

Percent visiting on survey day by income

0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%

Poor Middle Rich
Income quintile Percent
Metro Visiting Urban Visiting Rural Visiting

2.4 Travel by gender
It is beyond the scope of this work to look at
differences in travel by gender in any detail,
although many have argued that this is an
important issue in developing countries
(Grieco and Turner, 2003; Mashiri and
Mahapa, 2002). However, when considering
a basic analysis of the amount of travel by
gender an interesting trend emerges.
According to the survey fewer women
traveled on the survey day than men (6-8%
fewer). For the poor in urban and rural
areas, however, the degrees of travel are
similar with women traveling 3-4% less.

Percent traveling on survey day by income and gender

50.0%
55.0%
60.0%
65.0%
70.0%
75.0%
80.0%
85.0%
90.0%
95.0%
100.0%

Poor Middle Rich
Income quintile Percent
Metro Male Metro Female Rural Male Rural Female

Unpacking the reasons for these differences requires more analysis, and is discussed briefly
in the ‘further work’ section of this paper.

7
3. OVERALL VEHICULAR MODE USE
3.1 Introduction
This section discusses vehicle use by
the South African poor, and therefore
excludes discussion of the walking
trip, which is discussed later in the
paper. An NHTS question which
asked whether a vehicle had been
used at all in the previous seven days,
was analysed. Although this does not
give an indication of trip distance of
vehicle use, it gives an idea of overall
modal popularity. Considering the
metropolitan areas, it can be seen that
the poor, in common with the lowest
three income quintiles share a similar

Use of mode in seven days in metropolitan areas

by income

0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%

Poor Middle Rich
Income quintile Percent

Train Bus Minibus-taxi Car

vehicle use, with minibus taxi use between 32-36% and bus, car and train use between 4-
7% each. This is in contrast to the upper income group for whom car use is by far the most
popular (64%) while minibus taxi is 14%. This graph provides some reason for the
difference between the transport problem as generally perceived by the poor (which
concerns issues of minibus safety and facilities), and the rich (who are concerned with car
congestion).
The pattern is similar in urban areas, although train trips are not significant there, whilst in
rural areas bus use is more of a factor, and there is a smaller degree of car use in the
upper quintiles.
There are several vehicle types noted by the NHTS which were used only modestly in the
survey periods. However, some interesting trends can be seen when these are analysed
by income group. The use of metered taxis in metropolitan areas, although small in
percentage terms (1.8% overall), is highest in the poorest group (2.5%). One may
speculate that this is due to insufficient alternatives for emergency, difficult or late night
trips. In urban areas the use of metered taxis, and sedan taxis, by the poor is also higher
than average.
3.2 The minibus taxi

Problems and Issues of the poor with taxis
Top four metro/urban taxi problems for poor:
1. Safety from accidents (69%)
2. Facilities at taxi ranks (62%)
3. Roadworthiness of taxis (58%)
4. Behaviour of taxi drivers (57%)
(level of dissatisfaction in brackets)

Given the important role that the
minibus taxi plays in the life of the
poor, this has been analysed further,
using the attitudinal sections of the
NHTS. Notable are the high levels of
concern amongst the poor with issues
which probably require regulatory
control and strict enforcement before
improvement will be seen: safety,
roadworthiness and driver behaviour.

8
Also requiring public sector
investment are taxi facilities. The rural
poor, not surprisingly given they have,
in absolute terms less income than the
metropolitan poor, also express
dissatisfaction with taxi fares.
By contrast, both metro/urban and
rural poor express high levels of
satisfaction with travel time and
distance between home and taxi.

Top four rural taxi problems for poor:
1. Facilities at taxi ranks (69%)
2. Taxi fares (67%)
3. Safety from accidents (63%)
Roadworthiness of taxis (61%)
(level of dissatisfaction in brackets)
Top four acknowledgements for taxi service
from the metro/urban poor:
1. Travel time (79%)
2. Distance between home and taxi
(72%)
3. Frequency of taxis during the peak
(68%)
4. Waiting time (64%)
(level of satisfaction in brackets)
Top four acknowledgements for taxi service
from the rural poor:
1. Travel time (69%)
2. Distance between home and taxi
(57%)
3. Behaviour of taxi drivers (56%)
Security on walk to taxi (56%)
(level of satisfaction in brackets)

3.3 General public transport accessibility
Considering issues of general
accessibility to public transport, as
measured by percentage of
households within 15 minutes walk of a
station or stop, it is clear that in
metropolitan areas good public
transport access is income dependent,
except for the highest income brackets
which are either poorly served by
public transport or perceived to be.

Public transport accessibility: metropolitan area

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Train within 15 minutes Bus within 15 minutes
Taxi within 15 minutes

9
In urban areas access to all modes is
reasonably similar across income
groups, except in the highest income
group where taxi access is markedly
lower.

Public transport accessibility: urban area

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Train within 15 minutes Bus within 15 minutes
Taxi within 15 minutes

The reverse is true in rural areas,
where the poorest are better served by
bus and taxi than other income groups.
Train accessibility is very small across
all groups.

Public transport accessibility: rural area

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Train within 15 minutes Bus within 15 minutes
Taxi within 15 minutes

Actual walk times to public transport
stops have been calculates and plotted
according to income opposite. Three
issues emerge. Firstly the relatively
high access times in rural areas for
both bus and taxi. Secondly the trend
in metro/urban areas for a declining
time to public transport stops with
increasing income, except for the
highest income groups. This is in
contrast with rural areas which show a
different pattern.

Walking times to nearest bus stop

0
2
4
6
8
10
12
14
16
18
20

Poor Middle Rich

Income quintile

Mean time (minutes)

Metro Urban Rural

10
Thirdly, the metro areas have better
access times to bus than urban areas,
whilst in both areas access times to
taxis are broadly similar.

Walking times to nearest taxi stop

0
2
4
6
8
10
12
14
16
18
20

Poor Middle Rich

Income quintile

Mean time (minutes)

Metro Urban Rural

The central role played by taxis in
servicing the needs of the poor is
illustrated by plotting percentage of
households with no access. (Shown
as bars in figure opposite). In all areas
taxi service penetration for the poor is
relatively high, when compared with
train and bus. Few of the metro/urban
poor are unable to access any taxi
service at all. The train service
reaches the least numbers of people.
An interesting feature emerges when
the percentage of rich with no access
is also plotted. (Shown as lines in the

No access to pt service: poor and rich compared

0
10
20
30
40
50
60
70
80
90
100

Metro Urban Rural

Area

Percent with no service

0
10
20
30
40
50
60
70
80
90
100

Poor: no train service Poor: no bus service
Poor: no taxi service Rich: no train service
Rich: no bus service Rich: no taxi service
figure opposite). For train and bus, the rich are marginally better served (despite their
relatively low usage) whilst for taxi the picture is distinctly different. Far fewer poor have no
access than rich, which illustrates the role of the taxi as the mode of choice for the poor,
and the one which serves them best in terms of network accessibility.
3.4 Affordability
Considering issues of affordability, and
the spend on public transport, it is not
altogether surprising that the poor
spend the most in percentage terms on
public transport. This is clearly a
considerable burden, with almost 50%
of metro/urban households spending
more than 20% of their declared
income on public transport. Almost
80% of rural households spend more
than 20% of their income on public
transport.

Percent spending > 20% of income on PT

0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%

Poor Middle Rich

Income quintile

Percent spending >20% of income on PT

Total Metro Total Urban Total Rural

11
It could be argued that the bus
subsidies should, to an extent, be
assisting with this problem, but with
only 3-5% of the poor reporting bus
use during the survey period, (the
smallest reported use out of all the
income quintiles) this is evidently not a
major help in terms of affordability of
transport for the poor.

Percent of people reporting bus use in 7 days

0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%

Poor Middle Rich
Income quintile Percent

Metro
Urban
Rural

3.5 Vehicle ownership
Vehicle ownership by the poor is, not
surprisingly, very low. However, an
examination of the data shows that the
bicycle is the most commonly held
vehicle by the poor. Five percent of
metropolitan households; almost eight
percent of urban households; and over
seven percent of rural households own
at least one bicycle.

Vehicle ownership by poor

0
1
2
3
4
5
6
7
8

Bicycle (1)
Bicycle (2+)
Car (1)
Car (2+)
Commercial car
Minibus
Motorcycle

Vehicle type

Percent who own

Metro Urban Rural

12
Car ownership is a frequently quoted
statistic in transport planning work and
so the NHTS data on car ownership
was divided into the income quintiles
and compared with a selection of
international countries with similar
statistics (although the most recent
comprehensive dataset was 1998).
The comparison is very revealing,
indicating that the metro/urban rich
have car ownership broadly in line with
US/UK values, whilst the poor are
more in line with China, India and
Peru.

Car Ownership:

SA metropolitan (2003) and selected
international (1998)

United States
Italy

SA urban
middle

SA metro rich
United
Kingdom

SA metro poor

SA urban rich

Azerbaijan
SA metro
middle Peru
SA rural poor
India
China

Argentina
Korea, Rep
Greece
Israel

0
100
200
300
400
500
600

Cars/ 1000 population

4. ACCESSIBILITY TO EDUCATION

Mode to school for metropolitan poor

Walk
85%

Car
2.8%
Taxi
6%
Bus
4%
Other
1%
Train
1.1%

Mode to school for
urban poor

Walk
91%
Car
0.4%
Taxi
4%
Bus
2%
Other
3%
Train
0.0%

Mode to school for
rural poor

Walk
97%
Car
0.2%
Taxi
1%
Bus
1%
Other
1%
Train
0.0%

3.1 Mode split to education
It is clear from the pie charts of mode split to school that the school trip for the poor is
overwhelmingly a walking trip. This is especially so in rural areas where 97% of scholars
walk and only 1% use a bus or taxi. It can be argued that this is only problematic when

13

children face either unsafe or insecure journeys, or excessive walk times on their journey.
Walk times to education are considered below.
3.2 Time to education
Comparing travel time to income,
shown opposite, there is a general
trend of decreasing journey time to
school with increasing income, in
metropolitan areas. This may be
ascribed to the increase in vehicular
trips as incomes increase. The
poorest rural and urban children,
however, have a shorter journey time
than the next income group higher.
More analysis would be required to
identify the causes behind these
variations.

Average travel time to school by area

0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0

Poor Middle Rich

Income quintile

Time to school (minutes)

Metro Urban Rural

Time to school for walking metropolitan poor
1-15 mins
30%

16-30 mins
44%
31-45 mins
12%
46+ mins
14%

Time to school for walking
urban poor
46+ mins
9%
31-45 mins
9%

16-30 mins
36%

1-15 mins
46%

Time to school for walking
rural poor
46+ mins
19%

31-45 mins
15%

16-30 mins
32%
1-15 mins
34%

The issue of scholar fatigue due to long travel times to school may be examined by an
analysis of longest travel times to education. Depending on area 9-19% of poor scholars
face walk times of more than 46 minutes to and from school. Hence, in a typical rural
classroom of 40 poor children, 11 children will have walked for more than 31 minutes, and 2
of these for more than 61 minutes. However, these mean values mask extremes of walk
times, particularly in urban areas.

14
By plotting the standard deviation
(which is essentially a measure of
variation) with the mean of walk times,
it can be seen that the metropolitan
poor walk-times are more widely
varied for the poor than for other
income groups. Further analysis is
required to pinpoint these extremes
and address the inherent transport
and land use problems.
Transport planners wishing to address
problems of the education trip of the
poor are therefore faced with two
major issues : ensuring adequate
pedestrian infrastructure (well drained,
safe, secure and avoiding major
obstacles such as freeways or rivers);

Mean and standard deviation of walk time to
school in metropolitan areas

0.00
10.00
20.00
30.00
40.00
50.00
60.00

Poor Middle Rich

Income quintile

Time (minutes)

and liaising with education authorities in order to better serve students who are
compromised in their learning by long journeys to school.
5. ACCESSIBILITY TO LOCAL FOOD SHOPS
Given the importance of the shopping
trip to the poor, some further analysis
is worthwhile. Across all three areas
the mode use pattern is very similar,
with approximately 92% of households
walking to the nearest shop, with the
remainder mainly using taxis.

Mode use to nearest food shop by poor
Car/bakkie
2% Other
0%

Taxi
5%

Walk
93%

15
Access to shops in the metropolitan
and urban areas is good according to
the survey, with 85-90% having a food
shop within a 15 minute journey time.
An interesting feature emerges when
comparing less than 15 minute access
times to food shopping by income and
area, with the metropolitan poor
having slightly better access to the
nearest shop than the next two higher
income groups, despite the higher
income groups making more use of
taxi and car. One possible reason
could be linked to how ‘food shop’ is
defined by the different groups;

Percent within 15 minutes of food shop by income

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Metro Urban Rural
another reason could be the difference between informal areas (where food shops such as
spazzas will open according to market demand) and more formal areas which are regulated
by zoning systems. Analysis of the data according to household type would shed more light
on this issue.
6. ACCESSIBILITY TO HIGHER ORDER SHOPPING
Mode use to nearest post office by poor: metropolitan area
Other
3%
Do not need to
go
9%

Walk
43%

Taxi
40%
Bus
1%
Car/bakkie
4%

Mode use to nearest post office by poor:
urban area
Car/bakkie
2%
Bus
0%

Taxi
35%

Walk
56%

Do not need to
go
6%
Other
1%

Mode use to nearest post office by poor:
rural area
Other
4%
Do not need to
go
10%

Walk
35%

Taxi
41%
Bus
8%
Car/bakkie
2%

The NHTS survey does not distinguish between types of shopping, and so for the purposes
of this paper access to Post Offices has been used as a proxy for a higher order shopping
experience, since in many cases a Post Office will be located alongside a selection of other
services and shops.
When considering access to Post Offices, it can be seen that the walk mode is still the
majority mode for the poor in metro/urban areas, although in rural areas, perhaps due to
longer travel distances, the taxi is most popular. In metro/urban areas 40% of the poor can
reach a Post Office within 15 minutes travel time. Accessibility is seen to improve as
incomes increase.

16

Percentage with post office accessible
within 15 minutes

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Metro Urban Rural

Percentage who cannot get to post office

0
1
2
3
4
5
6
7
8
9
10

Poor Middle Rich

Income quintile

Percent

Metro Urban Rural
An interesting feature is the number of respondents who say they ‘can’t get there’. Only
0.5% of the urban poor say they cannot get to a Post Office whereas equivalent values for
rural areas were 3% and for metropolitan areas were 2%. For rural areas, one would expect
long distances to form a barrier to travel. In urban areas this would not be such a
consideration. In metropolitan areas distances may also be a barrier to accessibility, but
other barriers could also be involved: rail reserves, rivers, security estates, industrial areas,
and freeways.
7. ACCESSIBILITY TO MEDICAL FACILITIES
The access pattern to medical facilities follows the familiar pattern for the poor, with walking
being the majority mode in metro/urban areas, and with accessibility increasing with income.
Almost all poor respondents reported being able to access medical help but for 14-16% of
metro/urban dwellers this takes more than 31 minutes and in rural areas 33% need more
than 31 minutes to reach medical help.
Mode use to nearest medical service by poor: metropolitan area
Car/bakkie
4%
Bus
3%

Taxi
41%

Walk
50%

Can’t get there
1%
Other
1%

Mode use to nearest medical service by poor:
urban area
Other
1%
Can’t get there
1%

Walk
65%

Taxi
30%
Bus
1%
Car/bakkie
2%

Mode use to nearest medical service by poor:
rural area

Car/bakkie
2%
Bus
8%

Taxi
46%

Walk
42%

Can’t get there
1% Other
1%

17

Percentage with medical service within
15 minutes away by income

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Metro Urban Rural

Percentage with medical service more than
31 minutes away

0
10
20
30
40
50
60
70
80
90
100

Poor Middle Rich

Income quintile

Percent

Metro Urban Rural

8. WHAT ABOUT THE JOURNEY TO WORK?
For many transport planning exercises the journey to work in the morning peak is a major
focus, but the NHTS data shows that such a focus is mis-directed in a pro-poor planning
environment. An examination of employment figures by income shows why. In the poor
sector of the population, unemployment is a major social issue with unemployment amongst
the poorest quintile at approximately 80% in the metro/urban areas. In the rural areas this
rises to 93%. It is only amongst the richest quintiles of metro/urban areas that unemployment
is less than 40%. Whilst congestion in the morning peak is an issue for those concerned with
economic efficiency, it is evidently not a major priority issue for the poor.
Employment by income group: metropolitan areas

0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%

Poor Middle Rich

Income quintile

Percent employment

Employed: formal Employed: informal Not employed

Employment by income group: rural areas

0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%

Poor Middle Rich

Income quintile

Percent employment

Employed: formal Employed: informal Not employed

10. DISCUSSION
10.1 Major themes
This paper has highlighted a wide range of transport issues of the poor, using the 2003
National Household Transport Survey. In this final section some themes which have emerged
from the analysis work are highlighted, followed by a discussion of the need for further work
both on this survey data and into the future.
The first major theme is that of difference between the rural and metro/urban poverty
experiences. Whilst this is not entirely unexpected, the analysis highlights the need to treat
the poorest metro/urban dwellers differently from the poorest rural dwellers.

18

Secondly, there is a theme of overall better accessibility for the urban poor than for the
metropolitan poor. This is easy to explain, given the smaller urban footprint of the towns, and
the importance of the walking trip to the poorest, but it raises interesting questions about
future patterns of rural-urban migration. Given that there is better employment at present in
urban centres, and there is certainly better accessibility, it must surely be possible that
smaller urban centres may see higher inward migration than metropolitan centres in future.
The third theme has already been mentioned: the overarching importance to the poor of the
walking trip. This neglected mode of transport has been highly under-represented in planning
efforts, given its importance to those with least and walking infrastructure is not even
mentioned in many planning documents despite the daily discomfort faced by those with no
other options but to walk.
The fourth theme is an overall pattern of decreasing accessibility with decreasing income.
This was particularly true in metropolitan areas, and calls for an integrated planning approach
to public facilities especially, if it is to be redressed.
Finally the important role of the minibus taxi to the poor was clear throughout the analysis. As
a vehicular mode it dominated in the poor sector. In fact, it is only in the richest quintiles that
it does not dominate as the majority vehicular mode.
10.2 Further work
By necessity this paper, and the accompanying presentation, has given only a brief overview
of the results of the analysis. As for further work, it would be useful to contrast the approach
used, of discussing relative poverty through the use of income quintiles, by a more absolute
measure, such as an income cut-off point, or by a measure such as informally housed
households, in order to more directly assist in informal settlement planning. For local
applications of pro-poor transport planning a fuller examination of origins and destinations,
and data disaggregated by zones would be useful.
This analysis has only touched upon the issue of gender and transport. Grieco and Turner
have also worked for some years in the field of gender, poverty and transport and have made
some findings which suggest that to ignore the role of gender in poverty is a mistake (1997).
In particular the women in households are more likely to be involved in trip chaining. A typical
trip chain for a woman would be to accompany children to school, visit shop/s and perhaps an
institution such as a bank, then visit a relative, then head for home. Similarly, a woman is
more likely to combine a trip to work with a visiting or shopping trip. Given the complexity of
women’s travel behaviour, and that trip chaining has not been addressed directly by the

NHTS, it is possible that the NHTS may be under-reporting female travel, and mis-
representing their behaviour.

Another behaviour which is predominantly female is travel within constraints. As women are
most often the care-givers in a family unit their travel is often dependent on finding a
replacement care-giver, or on taking the child or relative on their journey. The implications of
this are two-fold, firstly any delays to a woman’s journey have knock-on consequences in
terms of lost time for a series of people; secondly a woman is more frequently burdened by
carrying young children. Thus long journey times, and journey distances, are a particular
burden to the female.

19

Finally, and particularly in the rural context, women and girls are most frequently head-loading
goods from shops, field and market. This feature of life is generally not subject to survey, and

the NHTS is no exception, although an examination of the shopping trip would imply head-
loading or carriage.

In conclusion, then, a gender-based analysis by income and purpose could identify some
interesting issues, and better inform planning efforts. Future NHTSs could consider this as a
further avenue of data collection although, admittedly, this would add complexity to an already
very comprehensive data collection exercise.
10.3 Conclusions: towards a more pro-poor transport planning effort
The mainstream poverty literature largely ignores the transport-poverty link, and transport
issues do not generally feature in high level policy statements. Nonetheless, there are a
group of academics who would argue that transport can contribute positively to poverty
reduction, and that it deserves more attention than has been traditionally the case. Gannon
and Liu of the World Bank recognized this in their discussion paper “Poverty and Transport”
(1997). More recently the World Bank have compiled “Cities on the move: A World Bank
urban transport strategy review (2002)” which links “urban and transport sector strategies with
a strong poverty focus”. The work of Vasconcellos has also been important, particularly at
raising issues of inequity in the transport systems of developing countries (2001).
The South African government has set the prerogative for a poverty reduction, and the NHTS
has enabled an analysis of transport problems associated with poverty. Given this work some
tentative conclusions can be drawn about what a pro-poor transport planning approach would
include.
9. TOWARDS PRO-POOR TRANSPORT PLANNING
There is the need for a two-pronged approach to transport planning in South Africa: firstly
‘economic’ transport planning, which addresses the needs of what Thabo Mbeki refers to as
the ‘First’ economy. This would concern itself with increasing the efficiency of the working
population, either through addressing work trip congestion, or business users of the transport
network, or with issues of freight. The beneficiaries of such planning would be the higher
income, employed population who have car ownership rates closer to a ‘developed’ UK/ US
rate, and, economists would argue, the economy as a whole would ultimately benefit.
Then there is transport planning for the so-called ‘Second economy’. This sector are
underemployed, or employed informally in the main. The policies in this sector need to be:

 enabling of upliftment: with a focus on access to education, job-seeking and job-
creation;

 ensuring adequate basic needs support with reasonable access to medical help, social
welfare, and food; and also on
 social support networks, which the NHTS has demonstrated to be a major trip purpose
amongst the poor. These informal networks are necessary to offset the economic
hardships faced by those with little income.

20

The social and economic transport planning of a developing country have different intentions
and so will inevitably be in conflict, but for too long the economic style planning imported from
the US/UK has dominated and a redress of emphasis is required if South Africa is to deliver
on its pro-poor policy agenda.
At a practical level pro-poor transport planning would:

 Provide safe, secure, direct, well maintained walk and walk/cycle paths, and road-
crossing facilities, especially to schools.

 Improve bus routing between informal areas and key services
 Partner with education authorities with respect to schools locations, not only in rural
areas but also in metros
 Partner with police and national government with respect to the regulation and
enforcement of improved safety standards for taxi vehicles and taxi driving
 Invest in the upgrading of taxi terminal facilities
Data collection and analysis efforts such as those initiated by the NHTS can only contribute
positively to the required new directions in South African transport planning.
REFERENCES
Department of Transport (DoT) (2005) Key results of the national household travel survey,
August 2005. Director General: Department of Transport, Pretoria. Available online at
http://www.transport.gov.za/projects/nts/keyresults.pdf
Gannon, C. and Liu, Z. (1997) Poverty and Transport. World Bank paper TWU-30.
September 1997. Available online at
http://www.worldbank.org/transport/pov&tra_resources.htm
Grieco, M. and Turner, J. (1997) Gender, poverty and transport: a call for policy attention.
Presentation notes of a talk delivered at the UN International Forum on Urban Poverty
(HABITAT). Florence, November, 1997. Available online at
http://www.geocities.com/transport_research/gender.htm
Mashiri, M. and Mahapa, S. (2003) Social exclusion and rural transport: a road improvement
project. Tshitwe, South Africa. In Fernando, P. and Porter, G. (2002) (Eds) Balancing the
Load: Women, gender and transport. Zed Books, London.
Mbeki, T. (2006) State of the Nation Address of the President of South Africa, Thabo Mbeki:
Joint Sitting of Parliament. 3 February 2006. Available online at
http://www.info.gov.za/speeches/2006/06020310531001.htm.
Noble, M., Ratcliffe, A. and Wright, G. (2004) Conceptualising, defining and measuring
poverty in South Africa: An argument for a consensual approach. Centre for the Analysis of
South African Social Policy. University of Oxford. Available online at
http://www.sarpn.org.za/documents/d0001015/index.php

21

Parnell, S. and Mosdell, T. (2003) Recognising, explaining and measuring chronic urban
poverty in South Africa. Conference paper presented at ‘Staying Poor: Chronic poverty and
development policy’, Institute for Development Policy and management, University of
Manchester, 7-9 April, 2003. Available online at
http://www.chronicpoverty.org/resources/conference_papers.html
Vasconcellos, E.A. (2001) Urban transport environment and equity: the case for developing
countries. Earthscan, London.
World Bank (2002) Cities on the move: A World Bank transport strategy review. Available
online at
http://www.worldbank.org/transport/urbtrans/cities_on_the_move.pdf
Other web references:
For the UK’s Make Poverty History campaign: http://www.makepovertyhistory.org/
For the UN Millennium Development Goals: http://www.un.org/millenniumgoals/
All websites accessed 8 March 2006.

Back To Top