Vulnerability and household welfare in Vietnam

Abstract

This thesis contains three studies which provide a complete set of vulnerability assessments in Vietnam.

The first study in the thesis estimates the extent of vulnerability and analyses who the vulnerable are. In addition, this study investigates the link between dynamic poverty and vulnerability, something which has rarely been done. To do this, the most common definition of vulnerability as ‘Vulnerability as Expected Poverty’ (VEP) is used, along with a data set extracted from three consecutive surveys from 2002 to 2006. The results reveal that, (i) vulnerability estimated using the reference line is more appropriate than when estimated using the actual poverty line for poverty prediction in the case of Vietnam; (ii) ex ante vulnerability in previous periods might translate to ex post poverty in the following periods though both vulnerability and the incidence of poverty tend to fall over time; (iii) the vulnerability of the poor may trap them in poverty; and (iv) the vulnerability of the non-poor could propel them into poverty.

The second study investigates sources of household vulnerability and responses to risks in rural Vietnam using data from Vietnam Access to Resources Household Surveys (VARHS). Vulnerability as low utility measure (VEU) is employed to estimate and distinguish the sources of vulnerability. Next the household’s behavior to cope with shocks is analyzed; and finally the effectiveness of the insurance mechanism is evaluated. The main findings are that: (i) the utility of the average household is 71 per cent less than the hypothetical situation without any risk or inequality in consumption, and idiosyncratic shocks contribute 50 per cent of the loss; (ii) households depend heavily on informal coping strategies such as food consumption reduction, savings withdrawal, taking children out of school, or capital depletion. The opportunity to borrow money from formal institutions is limited, while subsidies from the government or NGOs are available only in cases of natural disaster; and (iii) household consumption and income exhibit highly correlated variation, implying that existing informal insurance instruments are less effective than expected.

The third study provides new evidence on the impact of health insurance coverage on household vulnerability using Vietnam Access to Resources Household Surveys (VARHS) undertaken during 2010-2012. The outcomes of interest are the probability of falling into poverty (VEP) and the magnitude of utility loss (VEU). Since the data set is not from an intervention program, the propensity score- matching method is employed to construct treatment and control groups. Risk aversion is calculated and used as an important explanatory variable for health insurance enrollment. The implications of the study suggest actions for the government to attain its goal of universal health insurance coverage. The estimates show that health insurance coverage helps rural households in Vietnam reduce the idiosyncratic component of utility loss by 81 per cent and has the added benefit of reducing the probability of being poor by about 19 per cent. The reverse effect of the risk aversion on health insurance enrollment implies that not only is there a potential ‘rigidity’ effect on health insurance demand, but also that there are deficiencies in health insurance market.