World Food Policy Volume/Issue 2-2/3-1 Fall 2015/Spring 2016 | Page 95

World Food Policy

understanding the future risks also hold
for food insecurity. First, the definition of
households is essential. It has been found
that a wider definition is preferable in the
rural context as members of the household
might have migrated but still feel being
part of the household and thus sending
remittances to support the household’s
access to nutritious food over time. This
is particularly important for targeting
social protection programs (Gödecke
and Waibel 2015). Furthermore, Nguyen,
Raabe, and Grote (2015b) confirm that
migration is a livelihood support strategy
for rural households.

Second, the incorporation of a
module on shocks with information on
type of shock, shock severity, timing,
and coping strategies of households is
considered as being essential as it allows
us to capture the dynamics of food
security assessments. Shocks do affect
households with implications for their
food security. Thus, climate change and
HIV/AIDS, for example, are known to
have impact on food security (Rosegrant
and Cline 2013; Badolo and Kinda 2012).

Third, the time dimension is
critical, especially as shocks influence
the stability dimension of food security.
But also seasons have a significant
effect on food security with households
consistently worse off during monsoon
season as evidenced in Bangladesh
(Hillbruner and Egan 2008).

Many household panel datasets
already exist in developing and transition
countries (Baulch 2011). However, most
of them started in the early 1990s and
many of them have been discontinued
so that a huge gap exists in longer term
multi-purpose panel household surveys
in developing and emerging economies.

Furthermore, some of the existing
panels were not only of short duration,
but also had low sample sizes, a narrow
geographic focus, or were designed for
very specific purposes and thus did
not allow intercountry comparisons.
Alinovi, Mane, and Romano (2009)
admit that most papers on food security
are descriptive and lack panel data. They
mostly take an ex-post view because of
the multidimensionality of food security
models. Finally, none of these surveys are
focused on rural conditions where the
existence of poverty and food security
is widespread and persistent over long
periods of time.

Against the backdrop of the
existing panel databases in developing
countries, we present our long-term
panel project which tries to address
the above-mentioned shortcomings of
other panel datasets. Our long-term
panel project has been designed for food
security assessments as we will describe
more in detail in the next subsection.
Long-term panel project

The long-term panel project builds
on the structure and the collaborations
established by the Research Group DFG
FOR 756 funded by the German Research
Foundation. It continues and further
advances the existing high-quality panel
database from Thailand and Vietnam.
Until to date, household and village
surveys have been carried out in three
selected regions in Thailand and Vietnam
in 2007, 2008, 2010, and 2013, i.e., four
panel waves. In addition, in 2010, a
complementary migrant tracking survey
of 1,000 migrants in Ho Chi Minh City
and Bangkok was carried out. In 2011,
95