World Food Policy Volume/Issue 2-2/3-1 Fall 2015/Spring 2016 | Page 106
Income Diversity and Poverty Transitions: Evidence from Vietnam
Empirical Strategy
lowland, and coastal zones. The surveys
collect information on household
demographics,
health,
education,
Data
economic activities, employment, access
his study is based on panel to financial markets, public transfers,
household surveys from 2007, household expenditures, and assets, and
2008, and 2010 from the particularly on shocks and risks.
provinces of Hà Tĩnh (HT), Thừa Thiên
Huế (TTH), and Đắk Lắk (DL) in 1.1 The Drivers of Poverty Transitions
Vietnam for the purpose of the research
This study applies an MNL
project “Vulnerability in Southeast Asia”
model
presented
in Wooldridge (2002).
being run by a consortium of German
universities and local research institutes Changes in household poverty statuses
(see Klasen and Waibel 2012). The survey over a period can be classified into
covers more than 2000 households located several mutually exclusive outcomes. The
in rural and peri-urban areas in these MNL model determines the probability
three provinces. The three provinces have that household i experiences one of
a diversity of agricultural and ecological the mutually exclusive outcomes j. The
conditions with mountainous, highland, probability is expressed as
T
(
(1)
)
pij = P Yi = j =
e
β j xi
J
∑e
for j = 0,1,2,…, J ,
β k xi
k =1
where Yi is the outcome experienced by
household i, βk are the set of coefficients
to be estimated, and xi includes a
household’s covariates and their changes.
The model is, however, unidentified since
there is more than one solution for that
leads to the same probabilities Y = 0, Y =
(2)
(
)
pij = P Yi = j =
e
1, Y = 2,..., Y = J. To identify the model,
one of the βj must be set to zero, and all
other sets are estimated in relation to that
base category. For convenience, β0 is set
to zero; therefore, the above probability
function can be written as
β j xi
J
1+ ∑e
β k xi
(
)
, for j = 1, 2, … , J and pi 0 = P Yi = 0 =
1
J
1+ ∑e
k =1
k =1
106
β k xi