Newsletter

August 1999 - Number 4

The Indonesian Family Life Surveys (IFLS-1 and IFLS-2)

The Indonesian Family Life Survey (IFLS) is an ongoing project that collects multipurpose longitudinal survey data on individuals, households, families, communities, and facilities in Indonesia. The IFLS is unusual in that it is not only longitudinal but is also an integrated household and community survey. The project is a collaboration of RAND, UCLA, and the Lembaga Demografi, University of Indonesia. Conducted in 1993-1994, the first wave of the survey, IFLS-1, interviewed 7,224 people in 13 provinces across the country. The sample is representative of about 83 percent of the Indonesian population. The IFLS-1 data are available on our Web page (www.rand.org/FLS/IFLS).

Fielded four years later, in 1997­1998, the second wave, IFLS-2, sought to re-interview all the IFLS-1 households, and successfully recontacted over 94 percent of them. The IFLS-1 effort had interviewed the head, spouse, and a sub-sample of other household members. The IFLS-2 effort attempted to re-interview all of these people, as well as anyone who was age 26 or older in IFLS-1 (whether or not he or she had been administered an individual interview in 1993). It succeeded in interviewing over 91 percent of the target individuals. Many of the respondents had moved away from the place where they were interviewed in 1993; the effort followed all movers except those who had left Indonesia and those who had moved to a province that was not included in the IFLS sampling frame. IFLS-2 will be placed in the public domain by the end of 1999.

The Individual and Household Survey

The IFLS-1 household instrument collected a series of indicators of economic well-being at the household level, including expenditure, income, assets, transfers, and links with non-coresident family members. At the individual level, information was collected on migration, labor supply and earnings, wealth and nonlabor income, education, marriage, fertility and contraception, health status, and use of health services and family planning services. Many of these modules contain a comprehensive retrospective component.

The IFLS-2 instrument repeated most of the IFLS-1 instrument and added several sections that focus on, for example, decisionmaking within the household, participation in community activities, and behaviors during pregnancy. A key innovative feature of IFLS-2 is its inclusion of a series of physical-health assessments conducted by a nurse. The nurse measured each respondent's height, weight, blood pressure, pulse, lung capacity, and hemoglobin level. As a measure of physical functioning, respondents were timed as they rose from a sitting to a standing position. Additionally, respondents between the ages of 7 and 24 years were tested on their cognitive skills in Bahasa Indonesian and mathematics.

The Community and Facility Survey

The IFLS-1 included a detailed Community-Facility survey of infrastructure, prices, the availability and quality of health facilities and schools, and other community characteristics. The survey effort conducted interviews with community leaders and the head of the women's group. In addition, it sampled public health centers, private practitioners, community health posts, and elementary and junior and senior secondary schools, visiting as many as 12 facilities in each community and administering detailed questionnaires to respondents at those facilities.

The IFLS-2 Community-Facility survey builds on the IFLS-1 Community-Facility survey, but it offers a number of innovations. For example, the IFLS-2 survey includes an interview with a community elder about the traditional laws and customs that shape social behaviors in the community, helping to characterize the ethnic diversity of the IFLS communities, and uses GPS technology to obtain measures of the latitude and longitude of facilities, the community center, and the center of the Enumeration Area (EA) in which the IFLS households are located.

The Collapse of the Rupiah and IFLS-2+

IFLS-2 interviewing was completed just before the Indonesian rupiah collapsed. For this reason, it is a potentially valuable baseline for a follow-up survey to gain insights into the immediate effects of the crisis; some indications of the strategies individuals, families, and communities adopted in response to the crisis; and the likely medium-term and long-term effects of Indonesia's economic crisis. Time was of the essence. In July 1998, the survey team returned to the field and conducted IFLS-2+. Because neither time nor resources were available to mount a survey of the same magnitude as IFLS-2, the IFLS-2+ sample was smaller than the IFLS-2 sample, but most of the IFLS-2 instrument was retained. By design, IFLS-2+ re-administers many of the IFLS-1 and IFLS-2 questions so that comparisons across rounds can be made for characteristics of households and individuals, and for characteristics of communities and facilities, but to a 25-percent subsample of the EAs included in the fuller IFLS sample in 1993.

The IFLS-2+ sample was drawn in two stages. First, to reduce costs, the survey team revisited 7 of the 13 IFLS provinces, selecting provinces that best span the spectrum of socio-economic status and economic activity in the fuller IFLS sample. Second, within those provinces, the team purposely drew 80 EAs with weighted probabilities in order to match the IFLS sample as closely as possible. The IFLS-2+ sample is therefore representative of the entire IFLS sample. Counting all original households in IFLS-1 and the split-offs in IFLS-2--and including the 6 percent of households that were not found in IFLS-2--there are 2,066 households in the IFLS-2+ target sample. Over 95 percent of the target households were relocated and re-interviewed. If only those households found in IFLS-2 are considered, 98.5 percent of the target households were recontacted.

In addition to interviews with households and individuals, the IFLS-2+ effort repeated the IFLS-2 Community-Facility survey. Fieldworkers were instructed to re-interview both the community leaders and all the facilities interviewed in IFLS-2. For each community, interviewers were given a specific list of the names and addresses of the government health centers, private providers, community health posts, and schools from which data were collected in 1997. Of the providers interviewed in 1997, a total of 219 public providers (about 2.8 per community) and 387 private providers (about 4.8 per community) were re-interviewed in 1998.

Resources are currently being sought to support placing these data in the public domain.

Restructuring IFLS-1 for Compatibility with IFLS-2

As part of the IFLS-2 public release, RAND plans to reissue the IFLS-1 database in a form more compatible with the IFLS-2 database in order to facilitate the linking of these two databases for longitudinal analyses. Some data subfiles that are currently separate in IFLS-1 will be merged for ease of use and to parallel their IFLS-2 counterparts. To ensure linkage with IFLS-2 data for a given household or individual, a set of household and individual identifiers will be added to IFLS-1 files. For the Community-Facility data, a new set of facility identifiers that link to the IFLS-2 facility identifier coding scheme will be added. Some variables in IFLS-1 may be renamed to mirror their counterparts in IFLS-2. A "disposition" file will also be created and will tell users the status of a given IFLS-1 individual as of 1997. Some of the more obvious fixes in the IFLS-1 "fixes" programs will be applied to the revised IFLS-1 data. As well, adding recoded variables to IFLS-1 files will ensure consistent coding schemes between IFLS-1 and IFLS-2.