Living Standards Survey 2001
Timor-Leste, 2001
Get MicrodataIdentification
TLS_2001_TLSS_v01_M
Living Standards Survey 2001
| Name | Country code |
|---|---|
| Timor-Leste | TLS |
Living Standards Measurement Study [hh/lsms]
Timor-Leste experienced a fundamental social and economic upheaval after its people voted for independence from Indonesia in a referendum in August 1999. Population was displaced, and public and private infrastructure was destroyed or rendered inoperable. Soon after the violence ceased, the country began rebuilding itself with the support from UN agencies, the international donor community and NGOs. The government laid out a National Development Plan (NDP) with two central goals: to promote rapid, equitable and sustainable economic growth and to reduce poverty.
Formulating a national plan and poverty reduction strategy required data on poverty and living standards, and given the profound changes experienced, new data collection had to be undertaken to accurately assess the living conditions in the country. The Planning Commission of the Timor-Leste Transitional Authority undertook a Poverty Assessment Project along with the World Bank, the Asian Development Bank, the United Nations Development Programme and the Japanese International Cooperation Agency (JICA).
This project comprised three data collection activities on different aspects of living standards, which taken together, provide a comprehensive picture of well-being in Timor-Leste. The first component was the Suco Survey, which is a census of all 498 sucos (villages) in the country. It provides an inventory of existing social and physical infrastructure and of the economic characteristics of each suco, in addition to aldeia (hamlet) level population figures. It was carried out between February and April 2001.
A second element was the Timor-Leste Living Standards Measurement Survey (TLSS). This is a household survey with a nationally representative sample of 1,800 families from 100 sucos. It was designed to diagnose the extent, nature and causes of poverty, and to analyze policy options facing the country. It assembles comprehensive information on household demographics, housing and assets, household expenditures and some components of income, agriculture, labor market data, basic health and education, subjective perceptions of poverty and social capital.
Data collection was undertaken between end August and November 2001.
The final component was the Participatory Potential Assessment (PPA), which is a qualitative community survey in 48 aldeias in the 13 districts of the country to take stock of their assets, skills and strengths, identify the main challenges and priorities, and formulate strategies for tackling these within their communities. It was completed between November 2001 and January 2002.
Sample survey data [ssd]
Scope
1 Household information
A: Household Roster
B: New members since the violence in 1999
C: Persons leaving household after violence in1999
D: Information on parents of household members
2 Housing
A: Description of the dwelling
B: Housing state
C: Services
D: Ownership and expenditure
3 Access to facilities
4 Expenditures and consumption
A: Weekly food consumption
B: Monthly and annual non-food expenditure
C: Durable goods
5 Education
A: General education
B: Attendance school years 1998/99-2001/02
6 Health
A: Health care use
B: Chldren health
7 Fertility and maternity history
8 Employment
A: Labour force participation
B: Job information
C: Individual time use
9 Farming and livestock
A: Plots
B: Crops harvested
C: Agricultural inputs
D: Forestry
E: Farming equipment
F: Labour and farm produce
G: Livestock
H: Fishing and aquaculture
10 Transfers, borrowing and savings
A: Transfers given and loaned
B: Transfers received
C: Borrowing
D: Aid assistance
E: Savings
11 Other income
12 Social capital
13 Subjective wellbeing
A: Individual
B: Household
C: Vulnerability
Coverage
National coverage.
Domains: Urban/rural; Agro-ecological zones (Highlands, Lowlands, Western Region, Eastern Region, Central Region)
Producers and sponsors
| Name |
|---|
| National Statistics Directorate |
| Name | Role |
|---|---|
| The World Bank | Technical assistance |
Sampling
SAMPLE SIZE AND ANALYTIC DOMAINS
A survey relies on identifying a subgroup of a population that is representative both for the underlying population and for specific analytical domains of interest. The main objective of the TLSS is to derive a poverty profile for the country and salient population groups. The fundamental analytic domains identified are the Major Urban Centers (Dili and Baucau), the Other Urban Centers and the Rural Areas. The survey represents certain important sub-divisions of the Rural Areas, namely two major agro-ecologic zones (Lowlands and Highlands) and three broad geographic regions (West, Center and East). In addition to these domains, we can separate landlocked sucos (Inland) from those with sea access (Coast), and generate categories merging rural and urban strata along the geographic, altitude, and sea access dimensions. However, the TLSS does not provide detailed indicators for narrow geographic areas, such as postos or even districts. [Note: Timor-Leste is divided into 13 major units called districts. These are further subdivided into 67 postos (subdistricts), 498 sucos (villages) and 2,336 aldeias (sub-villages). The administrative structure is uniform throughout the country, including rural and urban areas.]
The survey has a sample size of 1,800 households, or about one percent of the total number of households in Timor-Leste. The experience of Living Standards Measurement Surveys in many countries - most of them substantially larger than Timor-Leste - has shown that samples of that size are sufficient for the requirements of a poverty assessment.
The survey domains were defined as follows. The Urban Area is divided into the Major Urban Centers (the 31 sucos in Dili and the 6 sucos in Baucau) and the Other Urban Centers (the remaining 34 urban sucos outside Dili and Baucau). The rest of the country (427 sucos in total) comprises the Rural Area. The grouping of sucos into urban and rural areas is based on the Indonesian classification. In addition, we separated rural sucos both by agro-ecological zones and geographic areas. With the help of the Geographic Information System developed at the Department of Agriculture, sucos were subsequently qualified as belonging to the Highlands or the Lowlands depending on the share of their surface above and below the 500 m level curve. The three westernmost districts (Oecussi, Bobonaro and Cova Lima) constitute the Western Region, the three easternmost districts (Baucau, Lautem and Viqueque) the Eastern Region, and the remaining seven districts (Aileu, Ainaro, Dili, Ermera, Liquica, Manufahi and Manatuto) belong to the Central Region.
SAMPLING STRATA AND SAMPLE ALLOCATION
Our next step was to ensure that each analytical domain contained a sufficient number of households. Assuming a uniform sampling fraction of approximately 1/100, a non-stratified 1,800-household sample would contain around 240 Major Urban households and 170 Other Urban households -too few to sustain representative and significant analyses. We therefore stratified the sample to separate the two urban areas from the rural areas. The rural strata were large enough so that its implicit stratification along agro-ecological and geographical dimensions was sufficient to ensure that these dimensions were represented proportionally to their share of the population. The final sample design by strata was as follows: 450 households in the Major Urban Centers (378 in Dili and 72 in Baucau), 252 households in the Other Urban Centers and 1,098 households in the Rural Areas.
SAMPLING STRATEGY
The sampling of households in each stratum, with the exception of Urban Dili, followed a 3-stage procedure. In the first stage, a certain number of sucos were selected with probability proportional to size (PPS). Hence 4 sucos were selected in Urban Baucau, 14 in Other Urban Centers and 61 in the Rural Areas. In the second stage, 3 aldeias in each suco were selected, again with probability proportional to size (PPS). In the third stage, 6 households were selected in each aldeia with equal probability (EP). This implies that the sample is approximately selfweighted within the stratum: all households in the stratum had the same chance of being visited
by the survey.
A simpler and more efficient 2-stage process was used for Urban Dili. In the first stage, 63 aldeias were selected with PPS and in the second stage 6 households with equal probability in each aldeia (for a total sample of 378 households). This procedure reduces sampling errors since the sample will be spread more than with the standard 3-stage process, but it can only be applied to Urban Dili as only there it was possible to sort the selected aldeias into groups of 3 aldeias located in close proximity of each other.
HOUSEHOLD LISTING
The final sampling stage requires choosing a certain number of households at random with equal probability in each of the aldeias selected by the previous sampling stages. This requires establishing the complete inventory of all households in these aldeias - a field task known as the household listing operation. The household listing operation also acquires importance as a benchmark for assessing the quality of the population data collected by the Suco Survey, which was conducted in February-March 2001. At that time, the number of households currently living in each aldeia was asked from the suco and aldeia chiefs, but there are reasons to suspect that these figures are biased. Specifically, certain suco and aldeia chiefs may have answered about households belonging, rather than currently living, in the aldeias, whereas others may have faced perverse incentives to report figures different from the actual ones. These biases are believed to be more serious in Dili than in the rest of the country.
Two operational approaches were considered for the household listing. One is the classical doorto-door (DTD) method that is generally used in most countries for this kind of operations. The second approach - which is specific of Timor-Leste - depends on the lists of families that are kept by most suco and aldeia chiefs in their offices. The prior-list-dependent (PLD) method is much faster, since it can be completed by a single enumerator in each aldeia, working most of the time in the premises of the suco or aldeia chief; however, it can be prone to biases depending on the accuracy and timeliness of the family lists.
After extensive empirical testing of the weaknesses and strengths of the two alternatives, we decided to use the DTD method in Dili and an improved version of the PLD method elsewhere. The improvements introduced to the PLD consisted in clarifying the concept of a household "currently living in the aldeia", both by intensive training and supervision of the enumerators and by making its meaning explicit in the form's wording (it means that the household members are regularly eating and sleeping in the aldeia at the time of the operation). In addition, the enumerators were asked to select a random sample of 10 households from the list, and visit them
physically to verify their presence and ask them a few questions.
Training for the listing operation was done on May 18 and 19, 2001 and was conducted by Manuel Mendonca, Juan Muñoz, Rodrigo Muñoz and Valerie Evans. It was stressed that it was important for the aldeia chiefs to understand that there was no aid coming as a result of this listing. The supervisors were also trained by Lourenco Soares and Rodrigo Muñoz to use the program installed on their laptops to record agricultural data being collected for JICA while the teams were in the field for the listing operation. This was an opportunity for the supervisors to become familiar with entering data in the field as a preparation for the TLSS. Finally, the listing
operation was carried out by 5 teams, each one comprising one supervisor and three enumerators, between May 21 and June 28.
See detailed information on selection probabilities and sampling weight calculations in document titled "Basic documentation".
Survey instrument
The 2001 TLSS household questionnaire follows the regular design of that of a Living Standards Measurement Study (LSMS) Survey. It was designed to collect all the necessary information required for a fairly comprehensive assessment of living standards and to provide the key indicators for social and economic planning. It comprises thirteen main sections and several subsections, each covering different topics about household activities. As a result, each household had to be visited at least two times to complete all sections.
Two additional sections are worth noticing when comparing this questionnaire with standard LSMS questionnaires. The first one refers to social capital, which tries to capture the involvement of the population in user or community groups and local networks as means of support for themselves both economic and socially. The second one is about subjective wellbeing. It covers individual perceptions on living standards, economically and power status and main concerns for the own individual and the country. It also provides information on consumption adequacy for food, housing, health, income, etc. Lastly, vulnerability, understood mainly as food insecurity, is addressed in this section too. Data are gathered on the number of months with inadequate food provision, members who suffered the most and coping strategies.
Data collection
| Start | End |
|---|---|
| 2001-08 | 2001-11 |
RECRUITMENT AND TRAINING
Part of the required workforce to carry out the survey fieldwork was drawn from the same teams that did the household listing. Indeed all of them were involved in this process too. This had the advantage that they knew already the location of the sucos and aldeias and had met their chiefs. Household listing records on how to access each aldeia, whether by vehicle or by foot, and the time to get there from the suco center had also been kept and were used for planning purposes. However, additional people were also recruited to complete the necessary teams for the fieldwork, specific language requirements were asked for most of them i.e. knowledge ofv
Fataluku, Bunak or Mambae. In the end, 37 people were trained and the best 32 were chosen for the enumeration. The best supervisor from the listing operation, Elias Dos Santos, was chosen to be the Field Coordinator and to assist in the enumerator training. The remaining 4 persons were kept as a backup and to do some work in Dili. Hence, eight field teams, each composed of three interviewers and one supervisor, conducted the household survey. Six teams were outside Dili, one for Oecussi and two in Dili, the main one and the spare team.
FIELDWORK
The survey was fielded during end August to early December 2001. Each team was responsible to cover one aldeia per week, so each interviewer had to interview 6 households during that period. Several visits to each household were required to complete all modules of the questionnaire.
Each of the 300 selected aldeias was to have 6 households interviewed for a total of 1,800 households. The questionnaires for each aldeia were sent out with a tracking sheet containing the names of the head of household for the 6 selected houses, and three reserve households in case the original households were not available. If an original household (numbered 1-6) was not interviewed, it was to be replaced with the first reserve household, numbered HH 7. If a second original household, or the first reserve, was not available, it was to be replaced with the second reserve household (HH 8), and so on for the third reserve household (HH9). For any
replacement, a full description of why the original household could not be interviewed was to be documented on the tracking sheet by the supervisors.
Overall, there were 303 cases were a household had to be replaced. Among the reasons given for non-completion of the interviews, a few points are interesting. The refusal rate was extremely low: there were only 6 refusals in the entire survey, and of those, only two were outright refusals. Second, there is a great deal of movement in the country and this constitutes the bulk of refusals, 255, although it must be said that most of them appear to be temporal movements. One reason why people leave temporarily their aldeia is because after the harvest they have to go somewhere else where they can find work, otherwise they have nothing to do and can not support themselves. The other explanation is that during planting time they have to move to their land for several weeks because that is at a considerable distance from their dwelling. Finally, the remaining 42 refusals were either because the dwelling could not be found or it was empty, or because the dwelling should not have been included on the listing.
Following completion of the fieldwork, a general debrief was held at the World Bank’s Dili offices with the participation of almost all supervisors and interviewers. The intention was to discuss issues and share experiences on the enumeration process such as their perceptions about their work, problems encountered, comments on sections of the questionnaire that were particularly hard to answer, level of cooperation of the chiefs and reception of the households interviewed. For instance, the health section seemed to be of special importance for the interviewees and many of them spoke about the need of more health services, the consumption module was considered a bit long, almost all women answered without major problems the fertility section, the Indonesian wording of some agricultural questions was ambiguous, chiefs were very cooperative and the participation of the households was more than satisfactory.
Data processing
A decentralized approach to data entry was adopted in Timor-Leste. Data entry proceeded side by side with data gathering with the help of laptops to ensure verification and correction in the field. The purpose of this procedure was twofold. First, it reduced the time of data processing because it was not necessary to send the questionnaires to the central office to be entered. More important, data were available for analysis very soon after the fieldwork was completed. And second, it allowed for immediate and extensive checks on data quality. Any inconsistency revealed at this stage was to be rectified by revisiting the households while still being in the village, and so, the need for later data editing was minimized. A second round of standard checks on data quality was also implemented in the project office in Dili upon retrieval of the data from the field teams. In general, with a few exceptions, the analysis has confirmed the high quality of the data entry and validation processes.
The data entry program was designed to check for data entry errors, coding mistakes, as well as to search for incomplete or inaccurate data collection. It was based upon two major types of checks. On the one hand, standard value-range checks were included. If the data entry operator entered data, which was outside the bounds of the programmed range, either because the number was not a pre-coded one or because it was extremely unlikely, the program would alert him. On the other hand, it also contained a series of checks to ensure that the data collected were internally consistent. The skip program used in the questionnaire was programmed into the data
entry software to ensure that the information entered was consistent to the desired skip pattern. For instance, if the code “3” was entered by mistake in a question where the only valid responses were “1” or “2”, the program would alert the operator. Similarly, if the household reported having purchased a particular good, the program would check to see if information on quantities and expenditure was also reported. However if the data entered into the computer matched the information provided in the questionnaires, the data entry operators were instructed not to make any changes to any of them. Such cases were brought to the attention of the supervisor, which
either corrected the mistake based on another information collected in the questionnaire or decided if a visit to that household was necessary.
Data Access
In receiving these data it is recognized that the data are supplied for use within your organization, and you agree to the following stipulations as conditions for the use of the data:
-
The data are supplied solely for the use described in this form and will not be made available to other organizations or individuals. Other organizations or individuals may request the data directly.
-
Three copies of all publications, conference papers, or other research reports based entirely or in part upon the requested data will be supplied to:
Manuel Mendonca
Director
National Statics Directorate
Caicoli, Dili, Timor Leste
e-mail: mmendonca@mopf.gov.tl
AND
The World Bank
Development Economics Research Group
LSMS Database Administrator
MSN MC3-306
1818 H Street, NW
Washington, DC 20433, USA
tel: (202) 473-9041
fax: (202) 522-1153
e-mail: lsms@worldbank.org
-
The researcher will refer to the 2001 Timor Leste Survey of Living Standards as the source of the information in all publications, conference papers, and manuscripts. At the same time, the National Statistics Directorate is not responsable for the estimations reported by the analyst(s).
-
Users who download the data may not pass the data to third parties.
-
The database cannot be used for commercial ends, nor can it be sold.
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
Contacts
| Name | Affiliation | URL | |
|---|---|---|---|
| LSMS Data Manager | The World Bank | lsms@worldbank.org | http://go.worldbank.org/QJVDZDKJ60 |
Metadata production
DDI_TLS_2001_TLSS_v01_M
| Name | Role |
|---|---|
| World Bank, Development Economics Data Group | Production of metadata |
2010-06-30
Metadata version
Version 02 (March 2011).