Data Collection

Data Collection

The data collection included two phases: qualitative field work and a household assets survey. In the qualitative phase, the primary methodology was Focus Group Discussions (FGDs), complemented by interviews with key informants and a compilation of the secondary literature. The focus groups focused on four themes: the accumulation of assets over the life cycle; the importance of assets; the market for assets; and household decision-making over asset acquisition and use.The quantitative phase of the study involved the planning and execution of household surveys. Data was collected on all physical and financial assets of households along with ownership information. The assets covered include housing, agricultural land, livestock, agricultural implements, non-farm economic activities and associated assets, consumer durables and financial assets. We also have data on household demographics, livelihoods, awareness of inheritance laws, recent shocks and coping strategies, decision-making and consumption expenditure. We interviewed two people within a household (the principal couple, whenever possible).

The Ecuador study “Assets, Gender and Poverty in Ecuador” is a collaborative endeavor between the Gender & Culture Program at the Latin American Faculty of Social Sciences-Ecuador (FLACSO-Ecuador) and the Center for Latin American Studies at the University of Florida, USA.

The study includes three phases: qualitative field work, a national household assets survey and analysis and dissemination. During the first phase, between August and December 2009, field work was carried out in three provinces of Ecuador: Pichincha and Azuay in the highlands, and Manabí on the coast (see Map 1). These provinces were chosen to be illustrative of different processes of development as well as socio-economic characteristics; factors posited to influence the possibilities for asset accumulation.

In each province, field work was carried out in at least three municipalities (known as cantones), always including the provincial capital as well as several predominantly rural municipalities. The rural municipalities were chosen to be illustrative of a range of income-generating activities that might facilitate women’s accumulation of assets: in Pichincha, the cut flower industry, the main source of rural wage employment for women in Ecuador; in Azuay, international migration and artisan production; and in Manabí, diverse agricultural activities alongside the fishing and tourism industries. The urban municipalities are all characterized by a wide variety of female income-generating activities in the formal and informal sectors; in addition, all have micro-credit and housing programs aimed at women.

The primary methodology was the focus group, complemented by interviews with key informants and a compilation of the secondary literature. All told, 40 focus groups were held, organized in collaboration with 23 organizations, ranging from women’s and peasant organizations and cooperatives to microcredit groups. Most of the groups consisted of all women, although at least one male-only rural group was organized in each province, and in some cases, several mixed-sex groups were held as well. Also, at least one focus group was organized with professional and businesswomen in each provincial capital. The focus groups focused on four themes: the accumulation of assets over the life cycle; the importance of assets; the market for assets; and household decision-making over asset acquisition and use. A total of 58 interviews were conducted with key informants, including lawyers, judges, notary publics, real estate agents, leaders of grassroots movements, NGO representatives and academics.

The second phase of the study, carried out between January and July 2010, consisted of the planning and execution of a nationally-representative household survey, known as the FLACSO-University of Florida Ecuador Household Assets Survey. The survey of 2,976 households was administered by the survey and public opinion consulting firm, HABITUS, S.A.

The sample frame was based on the 2001 National Population and Housing Census of Ecuador. A random, stratified sample of 250 census units was drawn. Stratification was based on HABITUS’s internal stratification of census units according to socio-economic level. The sample size was determined so that the survey be representative of two of Ecuador’s natural regions, the coast and highlands (representing 95% of Ecuador’s population), and of rural and urban areas. The cartographic household map was updated during the field work (except in dispersed rural areas) so that within each census unit the random selection of households reflect the unit’s actual composition in 2010. Following census definitions, rural areas were defined as those composed of 5,000 households or fewer.

The national survey was undertaken between April and June 2010. It was preceded by intensive training of HABITUS’s survey enumerators and a pilot survey of 150 households in the municipalities of Quito (Pichincha), Guayaquil (Guayas) and Portoviejo (Manabí).

The third phase of the study was carried out between August 2010 and June 2011, and included cleaning of the data and the creation of the appropriate household, individual and asset data sets; the generation of descriptive statistics on the key variables; the production of a pamphlet summarizing the key findings of the study; and public dissemination of findings. 

Data collection for Ghana had two components: the qualitative survey and the quantitative survey.

The Qualitative Survey

Data collection began in August 2009. The qualitative survey was comprised of Focus Group Discussions (FGDs) involving women and men, supplemented by key informant interviews. A total of eighteen FGDs of women and nine of men (covering nine out of ten administrative regions of the country) were carried out. The regions covered were the Northern, Upper East, Upper West, Eastern, Central, Western, Brong Ahafo, Greater Accra and Volta regions.

Five of the localities sampled for the qualitative survey are urban or peri-urban communities; these are in the Northern, Upper West, Brong Ahafo, Greater Accra and Western regions (see Table 1). The remaining four are rural communities in the Upper East, Eastern, Central and Volta regions. In terms of the ecological distribution of localities, three were in the northern savannah area, three in the coastal belt and three in the middle/forest belt of the country (Table 1). The communities were selected on the basis of ethnic composition and the inheritance regimes practiced in the community. All the communities selected from the northern savannah, Greater Accra and Volta regions practise the patrilineal system of inheritance; as opposed to the matrilineal system practised in the localities visited in the Eastern, Central, Western and Brong Ahafo regions. The only region that was not covered in the sample is the Ashanti region, which is located in the center of the country. The matrilineal inheritance system practiced in the Ashanti region is similar to that of the other regions practicing the matrilineal inheritance regime. In the urban communities the women were mainly traders, seamstresses and hairdressers. The rural women were mainly farmers, fish smokers or traders.

Table 1: Sampled Localities, Inheritance Regimes and Dominant Ethnic Group

Regions Community Rural/Urban Ecological Zone Inheritance Regime Dominant Ethnic Group
Northern Walelwale Urban Savannah Patrilineal Mamprusi
Upper West Wa Urban Savannah Patrilineal Wale
Upper East Basengo Rural Savannah Patrilineal Frafrai
Eastern Pokrom-Nsabaa Rural Middle/forest Matrilineal Akwapim
Central Saltpond-Ankaful Rural Coastal Matrilineal Fante
Western Esiamah Peri-urban Coastal/forest Matrilineal Nzema
Brong Ahafo Kintampo Urban Middle Matrilineal Mo
Greater Accra La Urban Coastal Patrilineal Ga-Dangbe
Volta Woeti Rural Middle/forest Patrilineal Ewe

Three FGDs (two women’s groups and one men’s group) were conducted in each of the localities sampled. The rationale behind the separation of women into two groups was to enable us to cover all the sets of questions or modules planned in a maximum of two hours. The FGDs covered four main themes: 

a) Livelihood, vulnerability and shocks; 
b) The meaning of and market for assets;
c) Life cycle analysis of asset ownership; and
d) Empowerment and decision-making.

The first two modules were administered to one women’s group and the remaining two modules were administered to the other women’s group in each community sampled. Two modules were administered in the men’s group in each community so that all the four modules were covered in two communities. The purpose of conducting FGDs for women and men was basically to obtain the perspectives of both sexes on the kinds of assets acquired by women and the mode or process of acquiring these assets. The questions posed to the key informants centered primarily on markets for assets and their prices, as well as the mode of inheritance and marital regimes.

Social amenities vary across the sampled communities. While the urban communities are endowed with education and health facilities, only one of the three rural communities visited had primary and junior high schools. Children in these localities attended school in nearby communities. Clinics in the nearby communities also provided health service for members of the community. All nine communities had access to mobile telephone communication. Three of the communities (two rural and one urban) did not have weekly markets. Roads leading to almost all the communities were in good shape, although a few were not tarred. All the communities were accessible year round. The main source of drinking water for the urban localities was pipe-borne water, while boreholes, streams and wells were the main source of drinking water for the rural communities. Five of the communities were reported to have NGOs operating in the community.

The Quantitative Survey

Households were randomly selected across the ten administrative regions of Ghana using a two-stage process. In the first stage, 143 enumeration areas were randomly selected. In the second stage, fifteen households were selected from each of the enumeration areas. The exception was the Upper East region, where 20 households were randomly selected from five of the enumeration areas. Table 2 below presents information on the distribution of enumeration areas and households across the ten administrative regions. The sample is not self-weighted. (Weights will be applied to the sample to make it nationally representative).

Table 2 : The Sample for the Quantitative Survey

Region Number of Enumeration Areas Number of Households % Distribution of Households
Western 15 225 10.37
Central 14 210 9.68
Greater Accra 17 255 11.75
Volta 14 210 9.68
Eastern 17 255 11.75
Ashanti 23 345 15.90
Brong Ahafo 17 255 11.75
Northern 15 225 10.37
Upper East 6 115 5.30
Upper West 5 75 3.46
       
Total 143 2170 100.00

In India, the study was restricted to the state of Karnataka. The data collection (qualitative and quantitative) was implemented by the market and social research agency, Sigma Research and Consulting Pvt. Ltd. The qualitative data collection was undertaken between August and October of 2009. 26 Focus Group Discussions (FGDs) were conducted across seven districts in the state, including the city of Bangalore. Each FGD covered on average between 10-12 participants and covered a wide spectrum of socio-economic groups in rural and urban areas (Table 1). The FGDs in the coastal districts also covered matrilineal communities.

 

Table 1: Distribution of FGDs

FGDs District Area Group Agency of Entry Into Community No. Participants
1. Bangalore Rural Rural SC/ST Men School Teacher 7
2. Bangalore Rural Rural General Women (higher economic group) Anganwadi (State Creche) Worker (AWW) 11
3. Bangalore Rural Rural SC/ST Women Municipal Councilor 11
4. Bangalore Rural Urban General Men Municipal Councilor 10
5. Chamarajanagar Rural SC/ST Men Gram Panchayat (GP) Secretary 8
6. Chamarajanagar Urban General Women AWW 10
7. Chamarajanagar Rural SC/ST Women Community Leader 9
8. Chamarajanagar Rural General Women School Teacher 8
9. Dakshina Kannada Urban General Women Municipal Councilor 8
10. Dakshina Kannada Rural SC/ST Men GP Member 9
11. Udupi Rural SC/ST Women GP Member 8
12. Udupi Rural General Men (higher economic group) GP President 10
13. Shimoga Urban General Women (higher economic group) AWW 10
14. Shimoga Rural General Men GP Member 10
15. Shimoga Rural SC/ST Women AWW 11
16. Shimoga Rural General Women GP President 10
17. Bellary Urban SC/ST Women AWW 9
18. Bellary Urban General Women Social Worker 8
19. Bellary Rural SC/ST Women AWW 10
20. Bellary Rural SC/ST Men GP Member 12
21. Bagalkote Urban SC/ST Women AWW 15
22. Bagalkote Urban General Men (higher economic group) GP Member 8
23. Bagalkote Rural SC/ST Women Social Worker 12
24. Bagalkote Rural General Women AWW 10
25. Bangalore City Metropolitan Urban General Men (higher economic group) Telephonic and personal recruitment 14
26. Bangalore City Metropolitan Urban General Men (higher economic group) Telephonic and personal recruitment 12

The quantitative data collection involved the survey of a state-representative sample of 4,800 households. Eight districts representing the four agro-climatic regions of the state were selected for the study. Some but not all districts covered in the qualitative section overlap with those covered for the quantitative section. 600 households were surveyed in each district, with two interviews in every household.

A stratified random sampling methodology was followed. Districts were randomly selected within each agro-climatic region, and taluks (sub-district administrative units) were selected using the Population Proportionate to Size (PPS) methodology. Primary Stage Sampling Units (PSUs) in the rural areas were villages and in urban areas were Electoral Booth areas. Villages were selected using the PPS methodology while electoral booths within a town were selected randomly. A mapping of the areas in all PSUs was undertaken and depending on the number of households in the PSU, either all of the households in the PSU were enumerated or some randomly selected areas were enumerated. The sample of households surveyed was randomly selected from the households thus enumerated. In every PSU, 25 households were surveyed.

 

Table 2: District Profiles

Districts

Female Literacy (%)

Women's Work Participation Rate (%)

Agricultural Wages for Women (Rs./day)

Proportion of Rural Population (%)

Total Geographical Area (hectares)

Human Development Index

Value

Rank in state

Value

Rank in state

Value

Rank in state

Value

Rank in state*

Value

Rank in state**

Value

Rank in state

Coastal

Dakshina Kannada

77.21

2

41.70

1

61.06

2

61.57

3

47,7149

6

0.722

2

Udupi

75.19

3

33.90

11

44.45

8

81.45

23

35,6446

2

0.714

3

Malnad

Shimoga

66.88

6

28.00

20

39.67

11

65.24

7

84,7784

22

0.673

5

Northern Maidan

Gadag

52.52

16

37.7

6

36.84

15

64.79

5

46,5715

5

0.634

13

Gulbarga

37.9

26

34.9

9

30.36

25

72.77

12

161,0208

27

0.564

26

Bidar

48.81

20

26.20

22

30.00

27

77.04

15

54,1765

9

0.599

21

Southern Maidan

Mysore

55.81

13

25.3

23

47.22

6

62.81

4

67,6382

16

0.631

14

Tumkur

56.94

12

41.3

2

35.00

19

80.38

20

106,4755

25

0.630

15

 

KARNATAKA

56.9

 

31.09

 

38.0

 

66.01

 

1,904,9836

 

0.65