World Food Policy Volume/Issue 2-2/3-1 Fall 2015/Spring 2016 | Page 96

Constructing a Database for Food Security Assessments in Southeast Asia

all panel households in the provinces
of Ubon Ratchathani (Thailand) and
Hué (Vietnam) were surveyed again
to capture medium-term agricultural
effects. An extension of the panel by
another six waves (until 2023) can make
this database truly unique in terms of the
length and continuity and the depth and
scope of the inquiries and thus allowing
to address many long-term development
questions including nutrition and food
insecurity.

In order to meet the objective of
conducting comprehensive food security
studies that can capture all four food
security dimensions, in principle two
questionnaires are needed: a village and
a comprehensive household survey. The
village questionnaire collects information
on the village characteristics, such as
its location and physical and social
infrastructure, and the socioeconomic
and demographic profile of the village
(e.g., age structure, welfare status of
households, unemployment rates, and
major income sources) strategies. The
household questionnaire includes a
range of modules which are shown in
Table 6. Although the questionnaire has
originally been developed for research on
vulnerability to poverty it can contribute
to food security studies through many
of its sections. Adding specific modules
on in-depth food security indicators
would be possible and has in fact been
implemented for the Cambodia and Laos
panel surveys. Also, the survey can be
complemented by selected focus group
discussions for example to determine
the Coping Strategies Index (CSI), as
suggested by Maxwell and Caldwell
(2008), and even randomized control
trials related to food security.

Our long-term panel project for
Thailand and Vietnam appears unique
and valuable in several respects (see
also Klasen and Waibel 2013). First, its
sampling design facilitates comparisons
of countries, regions, and provinces.
Second, the sample size at provincial
level is larger than any of the existing
panels and the period spanned will be
among the longest of any survey study.
Third, the database allows comparisons
of individual households and to some
extent also of individuals within
households. It consistently includes
migrants and, thus, covers an important
driver of household dynamics in
emerging market economies. Fourth, the
survey instrument contains all aspects of
the living standard measures and adds
comprehensive modules on shocks, risks,
expectations, and subjective assessments
of welfare. Fifth, it has a very low attrition
rate due to an incentive structure that
promotes ownership through active
participation of the scientists in the data
collection process and thus closely links
data collection with research output.

The long-term panel data
project will therefore provide unique
opportunities to advance scientific
research on poverty, vulnerability,
and food security. The value-added
of the project emanates both from its
long-term nature and the richness
of information that allow for a good
tracking of households’ expectations
and actions. It will permit the analysis
of long and medium term, dynamic, and
inter-temporal dimensions that is rarely
possible with existing datasets, including
the currently existing four waves of the
FOR756 panel. The panel datasets can
help to advance research related to food
96