2.2 Data and descriptive analysis
2.2.1 Data
This chapter employs data from the Vietnam Household Living Standard Sur-
vey 2012 (VHLSS2012) and the Vietnam Rural Urban Migration Survey 2013
(VRUM2013). While the VHLSS2012 is used to define urban households, the
VRUM2013 is used to define migrant households in the destination city.
The VHLSS is conducted by the Vietnam General Statistics Office (GSO)
every two years. The survey covers 64 provinces and 8 regions, and is repre-
sentative at national and regional levels in both rural and urban areas. The
VHLSS2012 consists of 9,399 households, 2,703 of which live in urban areas. The
survey provides detailed information on individual and household characteristics,
including income, expenditure, and household registration status. However, us-
ing the VHLSS to study the impact of the ho khau on migrant households is inadequate because the definition of migrant households in the VHLSS2012 is
less comprehensive than it is in the VRUM2013. Specifically, the VHLSS2012
does not include short-term and seasonal migrants, but only long-term migrants
who have already settled down in the destination city and are more likely to
obtain an urban ho khau. Therefore, we use the VRUMS2013, the most recent survey specifically designed for the study of rural–urban migration, to capture
Chapter 2. Consumption behavior of migrant households in Vietnam: Remittances, duration of stay and the household registration system
migrant households. This survey addresses the shortcomings of the VHLSS2012
by encompassing short-term and seasonal migrants.
The VRUM2013 is conducted by the Central Institute for Economic Manage-
ment of the Ministry of Planning and Investment, Vietnam under the technical
support from the Research School of Economics at the Australian National Uni-
versity. This survey collects data and information from 869 migrant households
and individuals from all rural areas in Vietnam who migrated to urban areas for
work purposes. The survey covers four major cities and provinces where there are
large rural-urban migration volumes, including Hanoi, Ho Chi Minh City, Binh
Duong, and Dong Nai. Hanoi is the capital city located in the north, and Ho
Chi Minh City the largest city and economic centre of Vietnam located in the
south. According to the 2009 Population and Housing Census data, Ho Chi Minh
City and Hanoi were the two most populous cities in the country. Binh Duong
and Dong Nai are two provinces adjacent to Ho Chi Minh City, and have become
highly industrialized through absorption of a heavy flow of rural–urban migrants
in recent years.
It is important to note that the four cities and provinces covered in the
VRUM2013 are at different administration levels. Hanoi is the capital city, Ho
Chi Minh is a first level city and Binh Duong and Dong Nai are two provinces
that have its own second level cities under its administration. These four cities
and provinces were selected for the VRUM2013 survey because they record the
largest volume of rural-urban migrants relative to other parts of the country.
Data from the 2009 Population and Housing Census indicate that net migration
rates (the difference between the number of immigrants (people coming into an
area) and the number of emigrants (people leaving the area) throughout the
year) are high at 116% for Ho Chi Minh City, 50% for Hanoi, 341.7% for Binh
Duong, and 68.4% for Dong Nai (Le et al., 2011). In recent years, 42% of all domestic migrants moved to the two largest cities, Ho Chi Minh City and Hanoi.
In addition, the variation in administration levels of the destination covered, the
VRUM2013 allows us to quantify the role of the household registration restrictions
in shaping consumption of migrants in the city because the restrictions imposed
by the household registration requirements differ across cities in Vietnam. They
are most stringent in Hanoi, followed by the other first level cities (Ho Chi Minh
City, Can Tho, Da Nang and Hai Phong).
In this chapter, we define urban households as those from the VHLSS2012
currently living in the cities where theirho khau is registered, and migrant house- holds as those from the VRUM13 without an urbanho khau at the current place of residence.7 It is important to note that these migrant households are selected
from the rural household base of VHLSS2012. This designate feature of the
VRUM2013 allows a comparison of the two surveys. Although the coverage of
the two surveys is different, we restrict the VHLSS2012 to households in the
four cities and provinces covered in VRUM2013, which are Hanoi, Ho Chi Minh
City, Binh Duong, and Dong Nai. After restricting the sample to those with
non-missing values in our selected variables, the analysis sample includes 1,256
households, with the migrant share amounting to 58% (730 households).
Household consumption is our key variable of interest and is defined as total
consumption spending in the destination city. For migrant households, this is
the amount consumed in the current location of residence, excluding remittances.
Due to the temporary nature of their stay in the city and close ties to family back
home, migrants are expected to consume less in the city. Further, the difference
in their consumption compared to urban households is expected to pertain mainly 7Following this definition, 123 migrant households from the VRUM2013 with an urban ho khau should be included in the urban household sample, having taken on the status of permanent migrant households. However, this inclusion may lead to a downward bias in the consumption gap between migrant and urban households. It is well-documented in the literature that although migrant households are relatively well-established after living in the city for an extended period and obtaining their urbanho khau, they still do not take on the behavior of urban households (Chen et al.,2012). For this reason, we exclude these 123 permanent migrant households from the main empirical analysis, but incorporate them later to check for robustness of the results.
Chapter 2. Consumption behavior of migrant households in Vietnam: Remittances, duration of stay and the household registration system
to non-food consumption. Therefore, we divide total consumption into food and
non-food consumption.
The two surveys measure consumption using different categories of spending.
The VHLSS2012 provides detailed information on expenditures for food, housing,
education, medical and health care, electricity/fuel and water, and communica-
tion and cultural activities. The VRUM2013 more simply divides household ex-
penditures into three broad categories: food, non-food and housing. The non-food
expenditure measure in the VRUM2013 absorbs all expenditure on education,
medical and health care, and communication and cultural activities. In addition,
although housing expenditure is collected in both surveys, the measure consists
of different components. Specifically, housing expenditure in the VHLSS2012
includes expenses on rent and home maintenance, while in the VRUM2013 it
contains rent, home maintenance, and spending on water, electricity, and fuel for
heating and cooking. To allow for comparison across the two surveys, all items
other than food had to be treated as non-food consumption. In other words, we
define household total consumption as the aggregation of food and non-food con-
sumption, in which non-food consumption contains all expenditures on housing,
education, medical and health care, electricity/fuel and water, and communica-
tion and cultural activities. The household income measure includes cash income
from two broad sources: wages and salaries, net revenue from farm and non-farm
economic activities. Both household income and consumption are measured in
Vietnamese dong (VND) in real terms.