THE SOUTH AFRICA I KNOW, THE HOME I UNDERSTAND

THE SOUTH AFRICA I KNOW, THE HOME I UNDERSTAND

Income &

Expenditure Su vey r 2010/11

Making our y mon y work fo e r

you

Income and Expenditure of Households 2010/2011

Statistical release

P0100

Private Bag X44, Pretoria, 0001, South Africa

User information service: +27(12) 310 8600, Fax: +27(12) 310 8500

Main switchboard: +27(12) 310 8911, Fax: +27(12) 321 7381

Website: www.statssa.gov.za, Email: info@statssa.gov.za

Statistical release

P0100

Income and Expenditure of Households 2010/2011

Embargoed until:

6 November 2012 10:30 am

Enquiries:

User Information Services

(012) 310 8600

Income and expenditure of households, 2010/2011 / Statistics South Africa Published by Statistics South Africa, Private Bag X44, Pretoria 0001

© Statistics South Africa, 2012

Users may apply or process this data, provided Statistics South Africa (Stats SA) is acknowledged as the original source of the data; that it is specified that the application and/or analysis is the result of the user's independent processing of the data; and that neither the basic data nor any reprocessed version or application thereof may be sold or offered for sale in any form whatsoever without prior permission from Stats SA.

Stats SA Library Cataloguing-in-Publication (CIP) Data

Income and expenditure of households, 2010/2011 / Statistics South Africa. Pretoria: Statistics South Africa, 2012

P0100 215pp

A complete set of Stats SA publications is available at Stats SA Library and the following libraries:

National Library of South Africa, Pretoria Division National Library of South Africa, Cape Town Division Library of Parliament, Cape Town

Bloemfontein Public Library

Natal Society Library, Pietermaritzburg Johannesburg Public Library

Eastern Cape Library Services, King William’s Town Central Regional Library, Polokwane

Central Reference Library, Nelspruit Central Reference Collection, Kimberley Central Reference Library, Mmabatho

This report is available on the Stats SA website: www.statssa.gov.za

Copies are obtainable from: Printing and Distribution, Statistics South Africa

Tel: (012) 310 8358 (012) 310 8093 (012) 310 8251

Email: inadp@statssa.gov.za

annelineb@statssa.gov.za

For technical enquiries please contact:

Werner Ruch

Email: wernerr@statssa.gov.za

Preface

This statistical release presents a selection of key findings and tables based on the data that were collected by Stats SA through the Income and Expenditure Survey (IES) 2010/2011 that was carried out during the period September 2010 to August 2011. This release focuses on the income, consumption and spending patterns of households at national and provincial levels.

PJ Lehohla

Statistician-General

Cautionary notes

Readers are cautioned to take the following into consideration:

Comparisons to Census 2011

• It is important to note that the estimates reflected by the IES 2010/2011 have not been weighted to the findings of the recently released Census 2011; instead, the survey has been weighted to mid-March 2011 population estimates. Although distributions in terms of proportions may be similar, there will be differences in the findings when it comes to actual numbers.

• The IES is a more detailed tool for measuring personal and household income and thus, direct one-to-one comparisons to Census 2011 income data is limited.

Rounding off

• Due to rounding, the displayed totals in the tables do not always match the sum of the displayed rows or

columns

Table of contents

1. Introduction... 1

1.1 Background of the survey ... 1

1.2 Outline of this report ... 2

2. Key findings ... 2

2.1 Results of the IES 2010/2011 ... 2

2.2 Comparison of IES 2010/2011 against IES 2005/2006 ... 3

3. Summary of the findings ... 6

3.1 Household composition ... 6

3.2 Household consumption expenditure ... 7

3.3 Household income ... 10

4. Findings ... 14

4.1 Food, beverages and tobacco ... 14

4.2 Clothing and footwear... 14

4.3 Housing, water, electricity, gas and other fuels ... 15

4.4 Furnishings, household equipment and routine maintenance of the dwelling ... 16

4.5 Health ... 17

4.6 Transport ... 18

4.7 Communication... 19

4.8 Recreation and culture... 20

4.9 Education... 21

4.10 Restaurants and hotels... 22

4.11 Miscellaneous goods and services ... 22

5. Explanatory notes ... 24

5.1 The instruments of data collection ... 24

5.2 How the IES 2010/2011 was conducted... 24

5.3 Time span ... 24

5.4 Response details ... 24

6. Technical notes ... 26

6.1 Sample... 26

6.2 Coverage ... 26

6.3 Data collection ... 26

6.4 Data processing... 26

6.5 Data editing and imputation ... 27

6.6 Data organisation... 27

6.7 Weighting... 28

6.8 Comparability... 30

7. Concepts and definitions ... 31

8. Tables... 33

Tables

Table 1: Main differences between IES 2000, IES 2005/2006 and IES 2010/2011 ... 1

Table 2: Total and average annual household consumption expenditure by main expenditure group ... 2

Table 3: Comparison of average annual household consumption expenditure by sex and population group of household head, as well as main expenditure groups ... 4

Table 4: Comparison of average annual household income by sex and population group of household head ... 5

Table 5: Percentage distribution of households by population group and sex of household head ... 6

Table 6: Annual household income by sex and population group of household head ... 11

Table 7: Percentage distribution of total household consumption expenditure on education attributed to public and private institutions by population group of household head ... 21

Table 8: Data collection activities by week for the IES 2010/2011 ... 24

Table 9: Response rates for the IES 2010/2011 ... 25

Table 10: Comparisons between the IES 2000, IES 2005/2006 and IES 2010/2011 ... 30

Figures

Figure 1: Percentage distribution of households by settlement type and sex of household

head ... 6 Figure 2: Average household size by settlement type and sex of household head ... 7 Figure 3: Percentage distribution of total annual household consumption expenditure by

main expenditure groups ... 7 Figure 4: Percentage distribution of annual household consumption expenditure on broad

expenditure groups by population group of household head ... 8 Figure 5: Average annual household consumption expenditure by population group of

household head ... 9 Figure 6: Percentage distribution of households by per capita expenditure quintiles and

population group of household head... 9 Figure 7: Percentage distribution of households by sex of household head and per capita

expenditure quintiles ... 10 Figure 8: Average annual household income by population group of household head... 11 Figure 9: Percentage distribution of households by per capita income quintiles and

population group of household head... 12 Figure 10: Percentage distribution of annual household income by main income groups and

sex of household head... 13 Figure 11: Proportion of total household consumption expenditure attributed to food,

beverages and tobacco by population group of household head ... 14 Figure 12: Proportion of total household consumption expenditure attributed to clothing and

footwear by province ... 15 Figure 13: Proportion of total household consumption expenditure attributed to housing,

water, electricity, gas and other fuels by population group of the household head... 16 Figure 14: Proportion of total household consumption expenditure attributed to furnishings,

household equipment and routine maintenance of the dwelling by sex of the

household head ... 17 Figure 15: Percentage distribution of household consumption expenditure on health in South

Africa by population group of household head... 18 Figure 16: Proportion of total household consumption expenditure attributed to transport by

population group of household head... 19 Figure 17: Proportion of total household consumption expenditure attributed to communication

by settlement type ... 20 Figure 18: Proportion of total household consumption expenditure attributed to recreation and

culture by province ... 20 Figure 19: Proportion of total household consumption expenditure attributed to education by

population group of household head... 21 Figure 20: Proportion of total household consumption expenditure attributed to restaurants

and hotels by population group of household head ... 22 Figure 21: Proportion of total household consumption expenditure attributed to miscellaneous

goods and services by population group of household head ... 23

1. Introduction

This report presents the results of the Income and Expenditure Survey (IES) conducted by Statistics South Africa (Stats SA) between September 2010 and August 2011. The primary objective of the IES is to provide relevant statistical information on household consumption expenditure patterns that will inform the updating of the consumer price index (CPI) basket of goods and services.

The information presented in this report was collected from 25 328 households across the country over a period of 12 months. The survey used a combination of the diary and recall methods. Households were required to complete their daily acquisitions in diaries provided by Stats SA for a period of two weeks and to answer a variety of questions from the household questionnaire administered by a Stats SA official over a four week period.

1.1 Background of the survey

The data collection methodology used for the IES 2010/2011, namely a combination of diary and recall methods, was first used in the IES 2005/2006. However, in an effort to improve diary reporting, the diary-keeping period was reduced from one month to two weeks. After extensive testing, the reduced diary-keeping showed an increase in the number of items reported in the weekly diary and had a noticeable impact on reducing respondent fatigue (meaning households were less likely to drop out during data collection). Despite the reduction to two weeks, the survey was still designed to ensure diary data for every day across the whole 12-month data collection period. In other words, where previously there were 12 survey periods (one per month), the IES 2010/2011 had 26 survey periods (each period covering exactly two weeks) that covered the 52 weeks of the year. Table 1 below illustrates the main differences between previous IESs.

Table 1: Main differences between IES 2000, IES 2005/2006 and IES 2010/2011

Distinguishing features IES 2000 IES 2005/2006 IES 2010/2011

Sample size 30 000 DUs 24 000 DUs 31 419 DUs

Methodology Recall Diary and recall Diary and recall

Household questionnaire One module Five modules Four modules

Diaries None Four weekly diaries Two weekly diaries

Goods Payment approach Acquisition approach Acquisition approach Services Payment approach Payment approach Payment approach Expenditure

data collection approach

Own production Consumption approach

Consumption approach

Consumption approach

Survey period Five weeks – Mainly

October 2000

One year – September 2005 to August 2006

One year – September 2010 to August 2011 Reference period: Food

expenditure September 2000 September 2005 to

August 2006

September 2010 to August 2011

Visits per household One Six Four

Classification of expenditure items Standard Trade

Classification COICOP COICOP

1.2 Outline of this report

This report has eight sections. This section (Section 1) provides the background information on the IES. The remaining sections are organised as follows:

• Section 2 presents key findings on consumption expenditure in South Africa;

• Section 3 provides a summary of the findings of the survey;

• Section 4 presents findings of the survey across the different consumption expenditure categories;

• Section 5 provides explanatory notes that will provide greater detail into the design and implementation of the survey;

• Section 6 focuses on technical issues relating to the survey;

• Section 7 provides detailed information on the relevant concepts and definitions; and

• Section 8 presents selected results in a series of tables.

2. Key findings

2.1 Results of the IES 2010/2011

The results of the Income and Expenditure Survey 2010/2011 show that the total annual household consumption expenditure between September 2010 and August 2011 is estimated at R1,25 trillion. The average South African household spent approximately R95 183 during the survey with the main components of that expenditure coming from housing, transport, food and miscellaneous goods and services.

Table 2: Total and average annual household consumption expenditure by main expenditure group Rand

Main expenditure groups Total (in millions)

Average

Percentage contribution

Food and non-alcoholic beverages 159 973 12 200 12,8

Alcoholic beverages and tobacco 13 697 1 045 1,1

Clothing and footwear 56 170 4 284 4,5

Housing, water, electricity, gas and other fuels 399 991 30 505 32,0

Furnishings, household equipment and routine

maintenance of the dwelling 63 944 4 877 5,1

Health 17 794 1 357 1,4

Transport 213 977 16 319 17,1

Communication 35 431 2 702 2,8

Recreation and culture 38 020 2 900 3,0

Education 33 355 2 544 2,7

Restaurants and hotels 30 332 2 313 2,4

Miscellaneous goods and services 183 614 14 003 14,7

Unclassified items 1 760 134 0,1

Total 1 248 058 95 183 100,0

* Due to rounding, figures do not necessarily add up to totals

Following the trend established by previous IESs, housing, water, electricity, gas and other fuels remains the largest contributor to household consumption expenditure across households. The group represents 32,0% of total household consumption expenditure, with the average household spending roughly R30 505 during the survey year. Contributing factors to this expenditure group include dramatic increases in electricity tariffs and a combination of higher estimation of house values (the base) relative to IES 2005/2006, and higher rental yields which are used to impute rental values for owner-occupied dwellings (proportion of the base).

When combined with consumption expenditure on food and non-alcoholic beverages (which is the fourth largest expenditure group by itself), the two groups account for approximately half of all consumption expenditure in the country. Essentially, one out of every two Rand spent by South African households goes towards housing and food.

Transport is the second largest expenditure group and is estimated at R214 billion or 17,1% of total household consumption expenditure. The average South African household spent approximately R16 319 on transport between September 2010 and August 2011.

Miscellaneous goods and services represents 14,7% of total annual household consumption expenditure, making it the third largest group. The reason for this high expenditure is rooted in the classification system used for the survey (COICOP), which includes insurance as part of the miscellaneous expenditure group. The by-product of this classification also results in what appears to be very low health expenditure (which is roughly 1,4% of total household consumption expenditure), but since medical aid contributions and medical insurance fall under miscellaneous expenditure, the health expenditure group just represents out-of-pocket expenditure by households.

On average, households spent approximately R1 357 on health.

Accounting for 12,8%, food and non-alcoholic beverages is the fourth largest contributor to household consumption expenditure. While the proportion may seem low, it is partly the result of higher housing and transport expenditure which pushes other proportions lower. On average, a typical household spent R12 200 on food and non-alcoholic beverages in the survey year.

Approximately 1,1% of consumption expenditure went to alcoholic beverages and tobacco, with an average expenditure of R1 045 on these items by households. However, due to the nature of diary surveys and compounded by the fact that these items are generally seen as sin expenditure, it is important to note that this figure is probably under-reported by the survey.

Furnishings, household equipment and routine maintenance of the dwelling during the time of the survey was measured at R64 billion or approximately 5,1% of total household consumption expenditure. This translates to an average expenditure of R4 877 per household during the survey.

The average South African household spent R4 284 on clothing and footwear between September 2010 and August 2011. In total, households spent around R56 billion on clothing and footwear, 4,5% of total consumption expenditure.

The other expenditure groups, namely recreation and culture (3,0%), communication (2,8%), education (2,7%) and restaurants and hotels (2,4%) accounted for the remaining R137 billion measured by the IES 2010/2011.

2.2 Comparison of IES 2010/2011 against IES 2005/2006

The table below presents the real change in consumption expenditure according to the sex and population group of

the household head, as well as the change by main expenditure groups. It indicates that average household

consumption expenditure in South Africa increased from R56 152 in 2005/2006 to approximately R95 183 in

2010/2011, which translates into a nominal increase of 69,5%. In constant 2011 prices, we see a real increase in

spending of 24,6% or roughly R18 779.

Table 3: Comparison of average annual household consumption expenditure by sex and population group of household head, as well as main expenditure groups

IES 2010/11

(R)

IES 2005/06 (R)

Real growth Increase in Rand terms

South Africa 95 183 76 404 24,6% 18 779

Sex of household head

Male 115 890 91 180 27,1% 24 711

Female 63 307 53 222 18,9% 10 085

Population group of household head

Black African 55 920 41 410 35,0% 14 510

Coloured 97 965 80 446 21,8% 17 519

Indian/Asian 198 695 141 252 40,7% 57 443

White 314 524 270 813 16,1% 43 711

Main expenditure groups

Food and non-alcoholic beverages 12 200 12 127 0,6% 73

Alcoholic beverages and tobacco 1 045 966 8,2% 79

Clothing and footwear 4 284 3 002 42,7% 1 282

Housing, water, electricity and other fuels 30 505 20 624 47,9% 9 881

Furnishings and household equipment 4 877 4 610 5,8% 267

Health 1 357 1 306 3,9% 51

Transport 16 319 13 396 21,8% 2 923

Communication 2 702 1 911 41,4% 791

Recreation and culture 2 900 2 928 -1,0% - 28

Education 2 544 2 019 26,0% 525

Restaurants and hotels 2 313 1 968 17,5% 345

Miscellaneous goods and services 14 003 11 311 23,8% 2 692

Unclassified items 134 236 -43,3% - 102

* Total inflation for all items used as a proxy for "unclassified items" price change

While all population groups made significant real gains in average annual consumption expenditure, Indian/Asian- headed households had the biggest gains both in Rand value terms (R57 443) and as a percentage increase (40,7%). Black African-headed households had the second largest percentage gain (35,0%), but the smallest increase in terms of Rand value (R14 510). Meanwhile, while white-headed households grew the least in percentage terms (16,1%), that translated into a R43 711 increase from 2005/2006. With regard to sex, both female and male-headed households made noteworthy gains, showing a real increase of 18,9% and 27,1%

respectively since 2005/2006. However, in Rand terms, the difference in gains was sizeable at R14 626 between male-headed households and their female counterparts.

Housing expenditure saw the largest increase between the two surveys, growing by 47,9% since 2005/2006. As has been noted, respondents were reporting relatively higher values for their houses compared to 2005/2006;

when combined with a higher rental yield used to calculate imputed rent on owner-occupied houses, this expenditure group shows the biggest gains, both in nominal and real terms. Nevertheless, there is a range of other reasons for such large gains in housing expenditure, such as the dramatic increases in electricity tariffs which increased more than three times that of general inflation (111,4% for electricity versus 37,3% for all items between 2005/2006 and 2010/2011).

After housing, the largest percentage increases in average annual household consumption expenditure were in

clothing and footwear (42,7%), communication (41,4%), education (26,0%), miscellaneous goods and services

(23,8%) and transport (21,8%). However, in Rand terms, clothing and footwear, transport and miscellaneous goods

and services contributed well over a third (36,7%) of the total increase (housing accounted for 52,6%). The growth

in these expenditure groups mirrors that of more developed countries and ultimately highlights the growth of a

black middle class in South Africa and a country that is slowly, but overall getting richer. Another common feature

(which we also see in South Africa) is a decline in the share of food expenditure, especially as higher incomes are used to pay for more expensive housing and transport.

Expenditure relating to recreation and culture (such as expenditure on hobbies, games, toys, etc.) shows the slight decline, with a real change of only -1,0% or roughly R28 less than 2005/2006. This is likely the result of residual lag from households that had to cut such items from their spending during the global financial crisis in 2008 and that have not yet made the decision to increase spending in this area again.

Change in health expenditure is tough to measure/interpret since all expenditure on medical aid and medical insurance is classified as miscellaneous expenditure according to COICOP. So the 3,9% increase since 2005/2006 simply reflects the increase in out-of-pocket expenditure on things like medicine, etc. Between 2005/2006 and 2010/11, items linked to health have seen higher than normal inflation (prices increased by 40,0% compared to the overall inflation of 37,3%). One can therefore infer that there have been no major shifts in how households deal with the financial burden of illness.

A very positive change that can be seen in these results has been the reduction in unclassified expenses. This highlights improvements made by the survey in data collection, data processing and data editing strategies and processes. Through these improvements, it is now possible to ensure that more items are linked to a corresponding COICOP code.

The 24,6% increase in spending by households in South Africa can be linked to various reasons, such as households incurring greater debt, the use of savings and/or increases to income. A comparison of income data between the two income and expenditure surveys reveals that the average annual household income experienced a real increase of 16,7% (approximately R17 141) since 2005/2006 (see Table 4 below). Thus, according to the findings of the survey, higher income is likely one of the main drivers for higher household consumption expenditure. Since respondents commonly under-report their income either through forgetfulness or out of misplaced concern that the information they report might be shared with the taxation authority, it is safe to assume that the income gains are actually larger than reported.

Table 4: Comparison of average annual household income by sex and population group of household head

IES 2010/11

(R)

IES 2005/06 (R)

Real Growth Increase in Rand terms

South Africa 119 542 102 401 16.7% 17 141

Sex of household head

Male 151 186 127 914 18.2% 23 272

Female 70 830 62 397 13.5% 8 433

Population group of household head

Black African 69 632 51 773 34.5% 17 859

Coloured 139 190 109 038 27.7% 30 152

Indian/Asian 252 724 184 711 36.8% 68 013

White 387 011 385 599 0.4% 1 412

Examining the increases to income according to population group, we see the largest increases happening in non- white households. Indian/Asian-headed households showed a 36,8% increase (an increase of R68 013), closely followed by black African-headed households at 34,5% (an increase of R17 859). Households headed by coloureds saw a 27,7% increase in income (an increase of R30 152), while white-headed households saw only a 0,4% real increase or roughly R1 412 more. Despite this significant growth in income in non-white households, there is still a tremendous gap between the population groups. White-headed households on average earn more than 5.5 times the income of the average black African-headed household. So while the income growth trend shows very positive signs, inequality remains a serious challenge for the country.

Although marginal, male-headed households are still seeing greater income growth than households headed by women. However, in most male-headed households, the income reported commonly includes a working wife, so it is possible that gains made by their wives give extra weight to the household’s overall income.

In conclusion, even though South Africans are getting richer, overall there are still very high levels of inequality

across and within population groups. It is important to note that while almost everyone reported making financial

gains between the two surveys, the majority of the population are doing so from a relatively low base (i.e. big

percentage gains do not necessarily translate into big Rand gains).

3. Summary of the findings 3.1 Household composition

Table 5: Percentage distribution of households by population group and sex of household head Sex of household head

Population group of household head

Total

(%) Male

(%)

Female (%)

Black African 76,6 57,1 42,9

Coloured 8,5 64,5 35,5

Indian/Asian 2,5 79,8 20,2

White 12,4 76,0 24,0

Total 100,0 60,6 39,4

Of the total number of households in South Africa, the survey found that slightly more than three quarters (76,6%) of the households were headed by black Africans. Approximately one in eight (12,4%) households were headed by whites with less than one in ten (8,5%) households headed by coloureds. Only 2,5% were headed by Indians/Asians.

The majority of households in South Africa are headed by men, with six out of every ten (60,6%) households being male-headed as opposed to four tenths (39,4%) being female-headed. While less than a quarter of white (24,0%) and Indian/Asian (20,2%) households were headed by women, this was the case for more than a third (35,5%) of coloured-headed households and more than two-fifths (42,9%) of black African-headed households.

Figure 1: Percentage distribution of households by settlement type and sex of household head

Figure 1 shows that three quarters (75,7%) of all households in rural formal areas were headed by men.

Approximately two-thirds of households in urban formal areas (65,7%) and urban informal areas (64,8%) were also

found to be headed by men. Female-headed households were in the majority in traditional areas, where more than

half (53,5%) of all households were headed by women.

The survey found the average household size to be 3,85. This average was found to vary across the different population groups with white-headed households showing an average household size of 2,86. For Indian/Asian- headed households the average size was 3,76, while black African-headed households had an average size of 3,98. Coloured-headed households were the largest with an average size of 4,06.

Figure 2: Average household size by settlement type and sex of household head

Figure 2 shows that for all households the average household size was higher in female-headed households (4,20) than male-headed households (3,62). This trend can be seen across all settlement types. The largest average household size was found to be amongst female-headed households in traditional areas at 4,99.

3.2 Household consumption expenditure

Figure 3: Percentage distribution of total annual household consumption expenditure by main expenditure groups

Expenditure on housing, water, electricity, gas and other fuels remains the largest contributor to total annual

household consumption expenditure in South Africa.

Figure 3 shows that approximately a third (32,0%) of total household expenditure was spent on this main expenditure group.

Transport was the second largest contributor of household expenditure, accounting for one in every six Rand (17,1%) spent by all households. Miscellaneous goods and services accounted for 14,7% of household expenditure, with insurance, financial services and personal care making up the bulk of this expenditure group. The other large expenditure group across all households was food, beverages (both alcoholic and non-alcoholic) and tobacco – slightly more than one out of every seven Rand (13,9%) was spent on this group.

Figure 3 shows the smaller expenditure groups to be those of furnishings, household equipment and maintenance (5,1%), clothing and footwear (4,5%), recreation and culture (3,0%), and so on. Expenditure on health accounted for only 1,4% of total annual household consumption expenditure. It is important to note that expenditure on medical aid schemes is classified as insurance and will therefore be a component of the miscellaneous goods and services expenditure group.

Figure 4: Percentage distribution of annual household consumption expenditure on broad expenditure groups by population group of household head

Figure 4 shows that almost a fifth (18,4%) of annual household consumption expenditure was spent on food and clothing. Black African households spent a far larger proportion on food and clothing than their coloured, Indian/Asian and white counterparts. Over a quarter (26,7%) of household expenditure for black African households was spent on food and clothing while in contrast, white households spent less than a tenth (9,2%) of their expenditure on food and clothing.

The proportion of consumption expenditure spent on health, education, recreation and culture was fairly constant across the different population groups. However, the proportion spent on housing, services, furnishings, equipment and maintenance increased from approximately a third amongst black African (31,9%) and coloured (34,7%) households to more than two-fifths amongst Indian/Asian (41,0%) and white (42,9%) households.

Transport and communication accounted for a fifth (20,0%) of all household expenditure. As can be seen in Figure 4, Indian/Asian households spent the largest proportion (23,7%) of their annual household consumption expenditure on this broad group.

1This main group includes the annual rental value of the dwelling unit or the annual estimated rental value of the dwelling unit if the dwelling unit was rented for free. In the case of owner-occupied dwelling units, an annual estimated rental value is calculated at roughly 7,7% (nationally) of the value of the dwelling unit (the rental yield may vary according to province).

Figure 5: Average annual household consumption expenditure by population group of household head

As Figure 5 shows, the survey found that the average South African household spent R95 183 during the period September 2010 to August 2011. The differences across the population groups were substantial. The average annual consumption expenditure for black African households was R55 920. This average was almost twice as large for coloured households at R97 965 per annum, with the average Indian/Asian household spending twice as much as coloured households at R198 695. The average expenditure for white households was the highest at R314 524 per annum.

The difference between the average annual consumption expenditure of male- and female-headed households was also large. The survey found female-headed households spent R63 307 on average per annum while male-headed households spent R115 890 on average per annum.

Figure 6: Percentage distribution of households by per capita expenditure quintiles and population group

of household head

Figure 6 shows the distribution of households according to per capita expenditure quintiles within each population group. The per capita expenditure quintiles have the following values:

• Upper quintile: R44 949 and above

• 4

quintile: R18 545 – R44 948

• 3

quintile: R10 010 – R18 544

• 2

quintile: R5 554 – R10 009

• Lower quintile: Up to R5 553

Almost half (48,7%) of black African households fell into the bottom two quintiles. This means that in approximately one out of every two black African households, each household was spending less than R10 010 per household member per annum (less than R835 per month or R28 per day). At the other end of the scale, less than one in ten (8,8%) black African households were in the upper quintile.

Figure 6 also shows that very few households amongst the Indian/Asian and white population groups fell into the bottom two quintiles. While a fifth (19,0%) of coloured households were found to be in the upper quintile, this was true of almost half (49,4%) of all Indian/Asian households and more than four-fifths (83,5%) of all white households.

Figure 7: Percentage distribution of households by sex of household head and per capita expenditure quintiles

Figure 7 shows the distribution of households within each quintile by the sex of the head of household. Earlier in the report it was mentioned that female-headed households had a lower average annual consumption expenditure than male-headed households. The distribution in Figure 7 further reflects that finding.

More than half (55,0%) of all households in the lower quintile were headed by women. The proportion of female- headed households decreases across each quintile, with only 47,0% of households in the second quintile and 38,6% of households in the third quintile headed by women. Less than a third (30,4%) of households in the fourth quintile and only a quarter (26,0%) of households in the upper quintile were headed by women.

3.3 Household income

The IES 2010/2011 estimated total household income in South Africa to be R1,57 trillion for the period between

September 2010 and August 2011.

Table 6: Annual household income by sex and population group of household head Head of household Rand amount

(in millions)

Proportion (%)

All households 1 567 455 100,0

Sex

Male 1 201 714 76,7

Female 365 741 23,3

Population group

Black African 699 154 44,6

Coloured 154 730 9,9

Indian/Asian 84 495 5,4

White 629 075 40,1

Table 6 shows that male-headed households accounted for more than three quarters (76,7%) of the total income while female-headed households earned less than a quarter (23,3%) of this income.

Black African households, which we have seen above accounting for more than three quarters of the total numbers of households in the country, earned less than half (44,6%) of the total annual household income. In contrast, white households make up only 12,4% of the total number of households, but earn two-fifths (40,1%) of the total income.

Coloured households (8,5% of the total number of households) earned 9,9% of the total income while Indian/Asian households (2,5% of all households) earned 5,4% of the total annual household income.

Figure 8: Average annual household income by population group of household head

As one would expect, the average annual household income by population group of the head of the household

followed a similar pattern to that of the average annual household consumption expenditure seen in Figure 5.

The survey found that the average household income across all households was R119 542 per annum. This average was noticeably lower for black African households at R69 632, while the average for coloured households was R139 190. Indian/Asian households had an average of R252 724 per annum and white households had an average of R387 011.

The average household income for male-headed households was R151 186, while for female-headed households this average was far lower at R70 830.

Figure 9: Percentage distribution of households by per capita income quintiles and population group of household head

Figure 9 shows the distribution of households according to per capita income quintiles within each population group. The per capita income quintiles have the following values:

• Upper quintile: R57 100 and above

• 4

quintile: R21 003 – R57 099

• 3

quintile: R9 887 – R21 002

• 2

quintile: R4 544 – R9 886

• Lower quintile: Up to R4 543

A quarter (24,7%) of all black African households fell within the lower quintile, while one in ten (10,1%) black African households were in the upper quintile. For coloured households, this distribution was reversed with one in ten (9,4%) coloured households falling in the lower quintile and one in five (20%) in the upper quintile.

In contrast, only 1,8% of white households were in the lower quintile as were 2,8% of Indian/ Asian households.

The vast majority (75,8%) of white households and more than two-fifths (43,4%) of Indian/Asian households were found in the upper quintile.

An analysis by sex of the household head revealed that more than half (54,4%) of all female-headed households

were in the bottom two quintiles, while almost half (48,8%) of all male-headed households were in the top two

quintiles.

Figure 10: Percentage distribution of annual household income by main income groups and sex of household head

Figure 10 indicates that the largest proportion of annual household income was derived from work, with 72,7% of all household income generated from this source. This was true for three-quarters (75,7%) of all income for male headed-households but only three-fifths (62,8%) of all income for female-headed households.

For female-headed households, pensions, social insurance and family allowances accounted for a tenth (10,9%) of all household income while a similar proportion (9,1%) came from income from individuals outside of the household or other income not elsewhere classified.

Income in the form of imputed rent on owned dwellings contributed 16,1% of all income for male-headed

households and 17,2% of all income for female-headed households.

4. Findings

4.1 Food, beverages and tobacco

Expenditure on food, beverages and tobacco accounted for 13,9% of total consumption expenditure in South Africa in the IES 2010/2011. However, there were differences between the population groups in terms of expenditure on food, beverages and tobacco. Black African households spent a fifth (19,9%) of their total consumption expenditure on this expenditure group, the highest across all population groups, while coloured households spent a similar proportion (18,6%). In contrast, Indian/Asian (7,4%) and white (7,2%) households spent less than a tenth of their expenditure on food, beverages and tobacco.

Figure 11: Proportion of total household consumption expenditure attributed to food, beverages and tobacco by population group of household head

At a provincial level, households in Limpopo (22,1%) and Eastern Cape (18,8%) spent the highest proportion of their household consumption expenditure on food, beverages and tobacco. Households in Gauteng spent only 9,5% of their household consumption expenditure on this expenditure group, constituting the lowest figure across all the provinces.

Households headed by females allocated a higher proportion of their household budget to food, beverages and tobacco compared to their male counterparts. The IES 2010/2011 results show that among female-headed households, almost a fifth (18,6%) of their total consumption expenditure was spent on food, beverages and tobacco compared to male-headed households where only 12,2% of their total consumption expenditure was spent on this expenditure group.

4.2 Clothing and footwear

According to the IES 2010/2011, on average, a South African household spent 4,5% of their total household consumption budget on clothing and footwear.

Figure 12 shows that at the provincial level, household consumption expenditure on clothing and footwear as a

proportion of total consumption expenditure was highest in Free State (6,3%) followed by Limpopo (5,9%) and

Mpumalanga (5,7%). Household consumption expenditure on clothing and footwear as a proportion was lowest in

Gauteng (3,5%) and Western Cape (4,1%).

Figure 12: Proportion of total household consumption expenditure attributed to clothing and footwear by province

Female-headed households allocated a higher proportion of their total household consumption expenditure to clothing and footwear than their male counterparts. The results of the survey show that only 4,0% of total household consumption expenditure of male-headed households went towards clothing and footwear while 5,8%

was spent by female-headed households.

Across the different population groups, black African households spent a higher proportion (6,8%) of their consumption expenditure on clothing and footwear relative to the proportions spent by coloured (5,1%), Indian/Asian (3,3%) and white (2,1%) households.

4.3 Housing, water, electricity, gas and other fuels

According to the survey, almost a third (32,0%) of all household consumption expenditure in South Africa is spent

on housing, water, electricity, gas and other fuels. This main group accounts for the highest proportion of

expenditure.

Figure 13: Proportion of total household consumption expenditure attributed to housing, water, electricity, gas and other fuels by population group of the household head

Figure 13 shows the proportion of household consumption expenditure on housing, water, electricity, gas and other fuels across the different population groups. Black African-headed households spent approximately a quarter (26,5%) and coloured-headed households spent three-tenths (30,4%) of their consumption expenditure on this group. Indian/Asian (36,9%) and white-headed (37,9%) households spent more than a third of their expenditure on housing, water, electricity, gas and other fuels.

The difference between male-headed households and female-headed households regarding this expenditure group was small. While male-headed households spent 32,4% of their expenditure on housing, water, electricity, gas and other fuels, this proportion was 31,0% for female-headed households.

4.4 Furnishings, household equipment and routine maintenance of the dwelling

Furnishings, household equipment and routine maintenance of the dwelling accounted for 5,1% of total household

consumption expenditure in South Africa. Female-headed households spent approximately 5,6% of their household

consumption expenditure on this group of expenditure items, which was slightly higher than the proportion spent by

male-headed households (4,9%) on this expenditure group.

Figure 14: Proportion of total household consumption expenditure attributed to furnishings, household equipment and routine maintenance of the dwelling by sex of the household head

Households in traditional areas spent the highest proportion (6,6%) on furnishings, household equipment and routine maintenance of the dwelling compared to households in urban formal areas which spent the lowest proportion (4,8%) of their total household consumption expenditure across the different settlement types on this expenditure group.

At provincial level, households in Free State (7,0%) and Limpopo (7,0%) spent the highest proportion of their total household consumption expenditure on furnishings, household equipment and routine maintenance of the dwelling compared to households in Gauteng which spent only 4,5% on this expenditure group.

4.5 Health

Expenditure on health contributed 1,4% to national household consumption expenditure. It is important to note that the health expenditure group excludes health insurance or medical aid subscriptions – these are covered in the insurance section under the miscellaneous goods and services expenditure group.

On average, male-headed and female-headed households spent the same proportion (1,4%) of their consumption

expenditure on health.

Figure 15: Percentage distribution of household consumption expenditure on health in South Africa by population group of household head

Pharmaceutical products was the major component of health expenditure, accounting for almost half (48,3%) of all household expenditure on health. This was true across all population groups, with black African households allocating the highest proportion (56,2%) on this component.

Medical services was another significant contributor to health expenditure, accounting for more than a third (36,4%) of the money spent by all households. As Figure 15 shows, Indian/Asian households spent more than two-fifths (42,8%) of their health expenditure on medical services.

While only 7,7% of all households' health expenditure was consumed by dental services, coloured households spent 17,6% of their health expenditure on these services.

4.6 Transport

Figure 16 shows that approximately one in every six Rand (17,1%) of household consumption expenditure went towards transport. A major component of transport expenditure was the purchase of vehicles (new and used), which alone was responsible for 7,9% of the country’s total household consumption expenditure.

The share spent on transport among male-headed households (18,4%) was larger than the corresponding

proportion for female-headed households (13,7%).

Figure 16: Proportion of total household consumption expenditure attributed to transport by population group of household head

The proportion of household consumption expenditure spent on transport by Indian/Asian-headed households was 20,5%. This was the highest among all population groups, followed by white (17,5%), black African (16,9%) and coloured-headed (14,7%) households.

4.7 Communication

Nationally, 2,8% of household consumption expenditure was spent on communication, the largest component of which was telephone and telefax services. There was some disparity across the provinces regarding the communication share of consumption expenditure – Limpopo had the lowest proportion at 2,0% and Western Cape the highest at 3,4%.

There was little disparity between population groups. White households spent 2,7% of their consumption

expenditure on communication (the lowest proportion) compared with 3,3% for Indians/Asian households (the

highest).

Figure 17: Proportion of total household consumption expenditure attributed to communication by settlement type

Figure 17 shows that urban informal households spent the highest share of their household consumption expenditure on communication (3,7%) whilst their urban counterparts in formal areas spent 2,9%. Rural formal households spent 3,1%, and traditional households only 2,4% on this consumption expenditure group.

4.8 Recreation and culture

Figure 18 shows that recreation and culture accounted for 3,0% of total household consumption expenditure nationally. The survey shows that Western Cape (4,0%) had the highest proportion of household consumption expenditure spent on recreation and culture, while the lowest proportions were spent in Limpopo (2,4%) and Eastern Cape (2,3%).

Figure 18: Proportion of total household consumption expenditure attributed to recreation and culture by

province

Male-headed households spent a higher proportion of their household consumption expenditure on recreation and culture (3,2%) than female-headed households (2,6%).

A similar proportion of household consumption expenditure was spent on recreation and culture by Indian/Asian (3,7%), white (3,6%) and coloured-headed (3,5%) households. Black African-headed households spent the lowest proportion (2,4%) on this consumption expenditure group.

4.9 Education

Figure 19 shows that education accounted for 2,7% of the total household consumption expenditure in South Africa. Education expenditure among Indian/Asian households accounted for 3,1 % of their total household consumption expenditure, followed by white households on 2,7%. Black African and coloured-headed households spent 2,6% of the household expenditure on education.

Figure 19: Proportion of total household consumption expenditure attributed to education by population group of household head

According to the survey, households headed by males allocated a slightly higher proportion of their consumption expenditure to education (2,7%) than female-headed households (2,5%).

As with health care, education in South Africa is provided by both the private and public sectors. Almost two in every three Rand (62,1%) spent by South African households on education went towards expenses in public institutions.

Table 7: Percentage distribution of total household consumption expenditure on education attributed to public and private institutions by population group of household head

Population group Public institution (%)

Private institution (%)

Black African 63,1 36,9

White 60,7 39,3

Indian/Asian 58,2 41,8

Coloured 66,4 33,6

Total 62,1 37,9

Indian/Asian-headed households spent the highest proportion (41,8%) of their education expenditure on private institutions, while coloured-headed households only spent a third (33,6%) of the education expenditure on private institutions.

4.10 Restaurants and hotels

The restaurants and hotels expenditure group accounted for 2,4% of household consumption expenditure according to the IES 2010/2011 results. Male-headed households spent a higher proportion (2,5%) of their budget on this group of expenditure items than female-headed households (2,2%).

Figure 20: Proportion of total household consumption expenditure attributed to restaurants and hotels by population group of household head

White-headed households spent 2,6% of their total household consumption expenditure on restaurant and hotel services compared to 2,4% for black African-headed households and 2,3% for coloured households. Indian/Asian- headed households spent the lowest proportion (1,9%) on this expenditure group.

Across the provinces, the highest proportion spent on restaurants and hotels was by households in Western Cape (2,9%) with the lowest proportion spent by those in Eastern Cape (1,6%).

4.11 Miscellaneous goods and services

This expenditure group consists of personal care, personal effects, social protection, insurance (including that

related to dwelling, health and transport), financial services and other services not classified elsewhere. Across all

households, one in every seven Rand (14,7%) was spent on this group.

Figure 21: Proportion of total household consumption expenditure attributed to miscellaneous goods and services by population group of household head

According to the survey, black African-headed households spent the smallest proportion of their expenditure on miscellaneous goods and services, allocating 12,9% of their expenditure on this group. While coloured-headed households spent 13,9% and Indian/Asian-headed households spent 14,7%, white-headed households spent the highest proportion on miscellaneous goods and services at 16,9%.

The survey also found that male-headed households spent a higher proportion (15,1%) of their total household

consumption expenditure on miscellaneous goods and services than female-headed households (13,7%).

5. Explanatory notes

5.1 The instruments of data collection

The Income and Expenditure Survey 2010/2011 used three data collection instruments, namely the household questionnaire, the weekly diary and the summary questionnaire.

5.1.1 Household questionnaire

The household questionnaire was a booklet of questions. The questions were administered to respondents during the course of the survey period. There were four modules in this questionnaire with eighteen subsections. The first module dealt with establishing the composition and structure of the household, as well as capturing particulars of all household members. The second, third and fourth modules collected information on different categories of consumption expenditure covering education, health, dwellings and services, clothing, footwear, expenditure when away from home, domestic workers, furniture and equipment, transport, computer and telecommunications.

Included in the fourth module is information on finance and banking, as well as particulars of income.

5.1.2 Weekly diaries

This is a booklet that was left with the responding household to track all acquisitions made by the household during the diary-keeping period. The household (after being trained by the Interviewer) was responsible for recording all their daily acquisitions as well as information about where they purchased the item (source) and the purpose of the item. A household completed a different diary for each week of the survey period.

5.1.3 Summary questionnaire

This is a booklet of questions that was for the sole use of the interviewer. The instrument had two primary functions. It served as a code list for interviewers when assigning codes for the classification of individual consumption according to purpose (COICOP) to reported items recorded in the weekly diary, and it also helped to summarise the household's total consumption expenditure on a weekly basis to allow the interviewers to better understand the household's acquisition patterns to ensure accuracy and completeness of the diary.

5.2 How the IES 2010/2011 was conducted

A household was in a sample for a period of four weeks. The survey instruments were administered in stages at different visits during the four weeks of data collection. A module was administered at the beginning of each week.

A detailed list of activities conducted each week is shown in the table below.

Table 8: Data collection activities by week for the IES 2010/2011 Week 0

(Week before the survey period)

Weeks 1 to 2 (The survey period)

Week 3

(Week after the survey period)

• Hand-over by publicity team

• Establish rapport with household

• Train household on diary completion

• Conduct interview 1

• Drop weekly diaries to be completed by household

• Conduct interviews 2 and 3

• Collect completed diaries for week 1

• Verify completed diaries for week 1

• Codification by means of the summary questionnaire

• Conduct interview 4

• Collect and verify completed diary for week 2

• Codification by means of the summary questionnaire

5.3 Time span

Data collection for the Income and Expenditure Survey 2010/2011 was conducted over a period of one year between September 2010 and August 2011.

5.4 Response details

From the 31 419 dwelling units sampled across South Africa, 33 420 households were identified. Out of these, there was a sample realisation of 27 665 (82,8%) households, with the remaining 5 755 (17,2%) households being classified as out of scope due to a number of reasons, such as listing error, vacant/unoccupied dwelling, etc.

The table below shows the response rates for the IES 2010/2011 according to province.

Table 9: Response rates for the IES 2010/2011 Province Response rate

(%)

RSA 91,6

Western Cape 94,3

Eastern Cape 95,8

Northern Cape 95,6

Free State 94,7

KwaZulu-Natal 92,3

North West 91,6

Gauteng 82,9

Mpumalanga 93,5

Limpopo 90,3

6. Technical notes

6.1 Sample

The sampling frame for the IES 2010/2011 was obtained from Statistics South Africa’s Master Sample (MS) based on the 2001 Population Census enumeration areas (EAs).

The scope of the Master Sample (MS) is national coverage of all households in South Africa and the target population consists of all qualifying persons and households in the country. The MS focuses on private dwelling units, workers’ hostels, residential hotels, nurses’ and doctors’ quarters, but excludes patients in hospitals or clinics, guests in hotels and guesthouses, prisoners in prisons, scholars and students in school or student hostels and the aged in old-age homes. In summary, it has been designed to cover all households living in private dwelling units and workers living in workers’ quarters in the country.

The IES 2010/2011 sample is based on an extended sample of 3 254 PSUs, which consists of the 3 080 PSUs in the Master Sample and a supplement of 174 urban PSUs selected from the PSU frame. The IES sample file contained 31 419 sampled dwelling units (DUs). The 31 419 sampled DUs consist of 31 007 DUs sampled from the 3 080 design PSUs in the Master Sample and 412 DUs from the supplemented 174 urban PSUs. In the case of multiple households at a sampled DU, all households in the DU were included.

6.2 Coverage

The IES 2010/2011 included all domestic households, holiday homes and all households in workers’ residences, such as mining hostels and dormitories for workers. It did not include institutions such as hospitals, prisons, old-age homes, student hostels and dormitories for scholars. Boarding houses, hotels, lodges and guesthouses were also excluded from the sample.

6.3 Data collection

There are three main approaches used to collect data on household consumption expenditure, namely the acquisition, the payment and the consumption approaches. All three methods were used at some stage during data collection for IES 2010/2011.

The acquisition approach entails taking into account the total value of goods and services acquired (not necessarily consumed, but for household consumption purposes) during a given period, whether the household paid for them or not. This is the general approach that was followed by the IES 2010/2011 for most of the items. Information on non-durable, semi-durable and durable items is collected using the acquisition approach.

The payment approach takes into account the total payment made for all goods and services in a given period, whether the household has started consuming them or not. This approach is followed when collecting data of expenditure on services such as education, health, insurance, etc.

The consumption approach takes into account the total value of all goods and services consumed or used during a given period. This approach is used when collecting information on own production.

6.4 Data processing

Data processing refers to a class of computer programmes that organise and manipulate large volumes of numeric data. Data processing involved the processing of completed instruments, i.e. household questionnaires, diaries and the summary questionnaires. Information received from these instruments collected during fieldwork was converted into electronic format represented by numbers or characters. The main method used for this conversion was scanning. To ensure quality electronic data, the data were verified as well as edited and checked for consistency according to the predetermined editing rules. All information contained in damaged instruments that could not be scanned was identified and transcribed onto clean instruments in order to be scanned.

6.4.1 Coding of acquired items

Coding is the process of assigning numerical values to responses to facilitate data capturing and processing in

general. The code list for acquired items was based on the United Nations’ Classification of Individual Consumption

According to Purpose (COICOP). Codes were assigned to expenditure items and listed in the diaries by field staff

while the head office staff did coding for the household questionnaire. During data processing, all assigned codes

were checked and improved when necessary to ensure and enhance quality.

6.5 Data editing and imputation

Decision table technology was used to develop edit rules which were compiled to produce SAS edit programmes for the purpose of identifying logical, inconsistency and out-of-range errors. Once identified, these errors were manually fixed on the system after verification against corresponding scanned images of the survey instruments.

The introduction of a 3-digit check number helped in the reduction of human error often introduced by COICOP coding in the field.

Within the money-metric sections of the household questionnaire, all data entries for specific expenditure items were considered for analysis even if the preceding filter question might have indicated otherwise. If there was conflict, the latter was changed to a ‘Yes’. All skips in the household questionnaire were treated in a similar fashion.

In the case of missing key demographic variables (age, sex and population group), hot-deck imputation was used.

Altogether, there were 5 persons (0,005%) whose sex was imputed and 10 persons (0,01%) for whom their population group was imputed.

Not all households that participated in the survey automatically qualified to be included in the final data.

Households were disqualified from inclusion if they did not have a completed household questionnaire or if the questionnaire did not have information for Section 1 (personal information) and/or Section 5 (housing). Households that did not have any diary information were also disqualified. Therefore, to be included in the final dataset, a household needed to have a household questionnaire with Sections 1 and 5 properly completed, and with diary information for at least one week.

If a household had only one week of diary information, imputation was done for the missing week. A random donor household was chosen from a cluster of households with similar characteristics and its corresponding diary was used to impute for the missing diary.

To deal with non-response on imputed rent for owner-occupied dwellings, rental yields, as published in the Rode and Associates Report from Quarters 1 and 2 of 2011 were used. Unlike in IES 2005/2006 where a single rental yield was applied across the board, a segmented approach was used in the application of rental yields, by type of housing and province. We applied an average rental yield per area based on the results of the Rode Report, except for Limpopo and North West, which were not covered by the Rode report. An average national rental yield was used for the two provinces.

6.6 Data organisation

Data collected from the IES 2010/2011 had to be reorganised to make sense to facilitate meaningful further analysis. Diary data were multiplied by 26 weeks to get an annual value. The questionnaire data were multiplied by 12 months if the COICOP (Classification of Individual Consumption by Purpose) was defined as monthly, and the questionnaire data were multiplied by 1 if the COICOP was defined as annual. This process is known as annualisation.

Since the survey took place over a period of twelve months (September 2010 to August 2011), it was necessary to benchmark the reported expenditure data to March 2011, which was midway into the survey year. All the expenditures incurred before the end of February 2011, i.e. before March 2011 were inflated to March 2011 prices and all the expenditures incurred after March 2011 were deflated back to March 2011 prices using the consumer price index (CPI) data.

Although the many data files were generated from the processed data, they all served as input files into what is considered the four primary data files that are released to the public, namely:

IES2010_persons – This is the persons file. This file contains all information regarding the persons in the households that responded to the survey. Particulars of each person as captured on the flap and in Section 2 of the household questionnaire are to be found in this file.

IES2010_households – This is the household file. This file has all information collected about household characteristics, including household assets and number of persons supported by the members of the household.

The information is extracted from Sections 1 to 15 of the household questionnaire.

IES2010_Total_ies – This file contains all the household expenditure and household income data. Total household expenditure data consist of all expenditure from the household questionnaire (Section 1 to Section 17) and the weekly diary. Household income data are extracted out of Section 18 of the household questionnaire. This file also has information on 'Area of Purchase' and 'Type of Retailer', which is collected in Section 5 to Section 16 of the household questionnaire and in forms 1 and 2 of the weekly diaries.

IES2010_Personsincome – This file contains information on personal income. Income at personal level is Section

18 of the household questionnaire.