The main objective of this survey is to provide a scientific framework
for a wide range of potential studies of the medium- and long-term
changes that are taking place during the process of development. The
survey is meant to remedy a major constraint on the understanding of
development in low-income countries - the absence of detailed,
multi-level and long-term scientific data that follows individuals over
time and describes both the natural and built environment in which the
individuals reside.
Our strategy is to permit the investigation of unexpected connections
between the multiple transformations that occur during the process of
economic development. To do so, we have implemented a large-scale,
nation-wide panel survey in Ghana that will extend for at least 15
years. The planned interval for re-surveying is three years. This
interval has been chosen to conserve on costs, because three years is
sufficient to begin to observe real changes, and to coordinate with the
other Yale EGC panel surveys taking place around the world.
Most data collection efforts are short-term - carried out a one point in
time; are limited in scope - collecting information on only a few
aspects of the lives of the persons in the study; and when there are
multiple rounds of data collection, individuals who leave the study area
are dropped. This latter means that the most mobile people are not
included in existing surveys and studies, perhaps substantially biasing
inferences about who benefits from and who bears the cost of the
development process. The design of this survey aims to mitigate the
selectivity associated with migration in assessing socio-economic
mobility, and will permit precise estimates from the randomization of
experimental interventions.
Technical Information
Sampling
The survey provides regionally representative data for the 10 regions of Ghana. In all, 5009 households from 334 Enumeration Areas (EAs) were interviewed. Fifteen households were selected from each of the EAs. The distribution of the enumeration areas across the regions in Ghana is presented below. The number of EAs for each region was proportionately allocated based on estimated 2009 population share for each region. EAs for Upper East and Upper West regions were over sampled to allow for a reasonable number of households to be interviewed in these regions
Region | # of EAs | Percent |
---|---|---|
Western | 31 | 9.3 |
Central | 28 | 8.4 |
Greater Accra | 39 | 11.7 |
Volta | 33 | 9.9 |
Eastern | 43 | 12.9 |
Ashanti | 60 | 18.0 |
Brong Ahafo | 34 | 10.2 |
Northern | 38 | 11.4 |
Upper East | 16 | 4.8 |
Upper West | 12 | 3.6 |
Total | 334 | 100 |
A two-stage stratified sample design was used for the survey.
Stratification was based on the regions of Ghana. The first stage
involved selecting geographical precincts, or clusters, from an updated
master sampling frame constructed from the 2000 Ghana Population and
Housing Census. A total of 334 clusters (census enumeration areas, or
EAs) were selected from the master sampling frame. The clusters were
randomly selected from the list of EAs in each region. The selection was
based a simple random technique.
The second stage of selection involved a listing of all the households
in the selected enumeration areas (clusters) and a simple random
sampling of 15 of the listed households from each selected cluster. The
primary objective of the second stage of selection was to ensure
adequate numbers of completed individual interviews to provide estimates
for key indicators with acceptable precision at the regional level.
Other sampling objectives were to facilitate manageable interviewer
workload within each sample area and to reduce the effects of
intra-class correlation within a sample area on the variance of the
survey estimates.
Since the design is not self-weighting, household sample weights have
been computed and applied for the estimation of the survey results in
this report. This was to facilitate estimation of the true contribution
of each selected cluster in the sample.
Questionnaires
The information gathered from the survey will assist decision makers in the formulation of economic and social policies to:
- Identify target groups for government assistance
- Construct models to stimulate the impact on individual groups of the various policy options and to analyze the impact of decisions that have already been implemented
- Access the economic situation on living conditions of households
- Provide benchmark data for district assemblies
To achieve these objectives, detailed data has been collected in the following subject areas:
- Demographic characteristics: employment, education, migration
- Information about non-resident spouses and relatives
- Assets:
- Household assets:
(i) Livestock (ii) Tools (iii) Durable Goods
- Financial assets:
(i) Borrowing (ii) Lending (iii) In-transfers (iv) Out-transfers (v) Savings
- Household assets:
- Agricultural Production
- Land information:
(i) Plot background (ii) Size (iii) Fallowing information, soil type, irrigation (iv) Investment, ownership, rental status (v) Crops (vi) Chemical inputs (vii) Tractor use (viii) Seeds (ix) Labour inputs
- Sales and storage:
(i) Revenues from crop production (ii) Crop stores
- Land information:
- Non-farm Household Enterprise
- Basic Information and Assets
(i) Basic information (ii) Enterprise assets
- Information about employees
(i) Information about all employees (ii) Information about four important employees (iii) Enterprises operating in the past 1 month (iv) Enterprise in a typical month
- Accounting: General enterprise
- Accounting: Trade/wholesale enterprise
- Accounting: Food enterprise
- Accounting: Services
- Basic Information and Assets
- Household Health
- Insurance
- Anthropometry
- Immunization
- Activities of daily living
- Miscellaneous health
- Health in the past 2 weeks
- Health in the past 12 month
- Womens' Health
- Fertility
- Power
- Mens' Health
- Reproductive Health
- Power
- Children's Module
- Young child health - children younger than 5 years old
- Digit span test- children aged 5-15
- Raven's Pattern Cognitive Assessment- children aged 5-15
- Math questions- children aged 9-26
- English questions- children aged 9-26
- Psychology/Social Networking
- Psychology
(i) Depression (ii) Subjective social welfare (iii) Regretted consumption (iv) Townsend questions (v) Trust and solidarity (vi) Time use
- Big 5 personality questions
- Social networking
- Information seeking
(i) Interaction with organizations (ii) Extension services (iii) Volunteerism
- Psychology
- Consumption Module
- Food items consumed
- Clothing and footwear
- Expenditure on other items in last 12 months
- Fuel and other lubricants
- Housing Characteristics
- Part A - Rent, water, light, cooking, waste disposal, building materials
- Part B - Dwelling type, ownership, living conditions, power supply, surroundings
Data Collection
The main field work for the survey covered a 6-month period (November 2009 to April 2010) in order to ensure that enough household baseline information was gathered. Seventeen teams were involved in the data collection. Each of the teams was made up of a Supervisor, a Senior Interviewer, four Interviewers and a Driver. A number of supervisory teams from ISSER visited the field at regular intervals to enforce good quality control for field operations and also shaped the direction for the survey. Due to the length and intensity of the survey, most households were surveyed over the course of multiple visits.
Data Processing
The processing of the survey data began shortly after the fieldwork
commenced. The first stage of data processing involved office editing
and post-coding. Questionnaires were edited to double-check for
completeness and consistency in the questionnaires returned, while the
post-coding served to define new response categories to pre-coded
question or define a response set for open ended questions. Once the
editing and post-coding were done, the questionnaires were passed on for
data entry.
The data entry program was designed in CSPro version 4.0. The entry
program was designed with the necessary skip patterns and consistency
checks to ensure adequate data quality and validity. All questionnaires
were entered twice (100 percent verification) and the two files were
compared for entry errors which were subsequently verified and corrected
with the questionnaires. The data entry was completed in August of 2010.
The consolidated data files in CSPro format were then converted to STATA
format for further consistency checks and cleaning.
Technical Documents
Data Access Policy
For the time being, access to the data is restricted to members of
the Yale and University of Ghana communities who are PhD candidates or
faculty. Interested researchers will be asked to provide information
about themselves and the use to which they will put the data.
If you are interested in using the data, please fill out the following
Google Form and you will receive a response shortly.
* Data from the Round 1 (baseline) survey will be made available
to researchers at other institutions upon the completion of the Round 2
survey, anticipated in 2013.
Contact Us
Please contact our research team at ghana.panel@yale.edu with to share any comments, questions, or concerns. We look forward to hearing from you!