Research Projects

Study of the Tsunami Aftermath and Recovery (STAR)
Mexican Family Life Survey (MxFLS)
Work and Iron Status Evaluation (WISE)
Indonesia Family Life Survey (IFLS)

On December 26, 2004, a massive earthquake that registered 9.2 on the Richter Scale struck in the Indian Ocean off the coast of Aceh, Indonesia, spawning a tsunami that slammed into the island of Sumatra and caused unprecedented death and devastation. The tsunami subsequently traveled across the Indian Ocean wreaking havoc on coastal areas throughout the region. Worldwide, casualties are thought to be around a quarter of a million people. In Aceh, about 160,000 people died and half a million people were displaced. The largest international relief and reconstruction effort was mounted in the affected countries and in Indonesia there was a major effort to "Build Back Better."

STAR is designed to provide scientific evidence about the immediate and longer-term impacts of the earthquake and tsunami on the health and well-being of the Acehenese and North Sumatrans living along the coast before the earthquake. A key feature of the study is that we use population-representative longitudinal survey data collected both before and after the disaster. The study documents the immediate consequences of the disaster for mortality, family disruption and relocation, physical and psycho-social health, including biomarkers that measure cardio-metabolic risks, cognition, demogaphic behaviors and markers of social and economic well-being By re-interviewing the same respondents annually for five years after the tsunami, and again 10 and 15 years after the tsunami, the project provides evidence on the longer-term consequences of the tsunami on a broad array of indicators of well-being. Further, using deep learning methods, we are developing measures of destruction and reconstruction of the built and natural environment extracted from remote-sensed satellite imagery.

Analyses of these longitudinal data trace the reconstruction of lives and livelihoods in the aftermath of the disaster, paying particular attention to the roles of own and family resources, kinship and social networks, community strength, and receipt and leveraging of external aid. In so doing, the research provides characterizations of the individuals, households and communities that are associated with resilience to the deleterious consequences of this large scale exogenous shock.

STAR is co-directed by Elizabeth Frankenberg and Thomas. The project is funded by the National Institutes of Health (NIA and NICHD) the National Science Foundation, the World Bank, the Wellcome Trust, MacArthur Foundation and Hewlett Foundation.

  For a brief overview of key results, click here.
  For papers based on STAR data, click here.
  For more information about the project, click here.

Follow-up surveys were conducted in 2005-6 and 2009-13. About 90% of surviving baseline respondents were re-interviewed in the first follow-up; of the respondents who are thought to have moved to the United States, 91% were interviewed. In the second follow-up, about 85% of surviving baseline respondents were re-interviewed; among those who are thought to have lived in United States since baseline, 85% have been re-interviewed either in the United States or in Mexico if the respondent has returned. This is the first population-representative study of Mexican families with members living in both Mexico and the United States.

The survey collects extensive information on the social, economic and demographic characteristics of respondents, their households, their families and their communities. In addition to colelction a broad array of self-reported indicators of health, anthropometry, blood pressure, cholesterol, hemoglobin, glycosolated hemoglobin and C reactive protein are measured.

For more information, click here.

WISE is a large scale field experiment designed to identify the causal effect of an improvement in the health of an individual on the economic and social prosperity of that individual, his/her family and ecommunity. The double-blind randomized controlled trial was conducted in Central Java, Indonesia. There are high levels of iron deficiency in the population, particularly among older adults. Study subjects were randomly assigned to receive an iron supplement or a placebo on a weekly basis for a year. intervention in a study site in Central Java, Indonesia.

Beginning in 2002, over 12,000 individual members of more than 4,000 households were interviewed every four months for 4 years. The first two waves of the survey served as the baselines. After the second wave, households were randomly assigned to one of two groups. A supply of iron supplements was provided to each members of households assigned to one group. A supply of identical looking placeboes was provided to each members of households assigned to the other group. We asked participants to take the supplement/placebo on a weekly basis for a year. All participants who were supplemented would be expected to be iron replete by the end of the year.

Data were collected at the individual, household and community level. The health of each individual, including levels of iron in the blood (hemoglobin and transferrin receptors) was measured every four months. Individual level data covered time allocation including work and participation in community activities, earnings, self-reported and physical health and cognition. All adult household members were interviewed; questions about children were answered by a caretaker. Data on wealth and consumption were collected at the household level. Community resources, services and infrastructure were gathered along with detailed price data.

The project is a collaboration with Elizabeth Frankenberg, Jed Friedman, Jean Pierre Habicht, Mohammed Hakimi, Nick Ingwersen, Gretel Pelto, Teresa Seeman, Bondan Sikoki, James Smith, Cecep Sumantri, Wayan Suriastini and Siswanto Wilopo. It is funded by the National Institutes of Health (Fogarty International Center and NIA).

The study is led by John Strauss with Bondan Sikoki and Firman Witoelar. Public use data are available here.