Portal de Periódicos da CAPES

Vulnerability and poverty dynamics in Vietnam

2010; Taylor & Francis; Volume: 43; Issue: 25 Linguagem: Inglês

10.1080/00036841003670754

1466-4283

Katsushi S. Imai, Raghav Gaiha, Woojin Kang,

Income, Poverty, and Inequality

Abstract Drawing upon the Vietnam Household Living Standards Survey (VHLSS) data that cover the whole of Vietnam in 2002 and 2004, ex ante measures of vulnerability are constructed. These are then compared with static indicators of poverty (i.e. the headcount ratio in a particular year). Detailed analyses of the panel data show that (i) in general, vulnerability in 2002 translates into poverty in 2004; (ii) vulnerability of the poor tends to perpetuate their poverty and (iii) sections of the nonpoor but vulnerable slip into poverty. Durable reduction in poverty is conditional on (i) accurate identification of the vulnerable, (ii) their sources of vulnerability and (iii) design of social safety nets that would enable the vulnerable to reduce risks and cope better with rapid integration of markets with the larger global economy. Notes 1 However, it should be noted that there was considerable regional diversity in collectivization. For example, the South was never collectivized in any form (see, e.g. Kees van Donge, Citation2003). 2 Poverty rates used here are based on the international poverty line, which was devised by the Vietnamese GSO to reflect food expenditure for an intake of 2100 calories a day and corresponding nonfood expenditure. The basket of food and nonfood items is determined by the consumption patterns of the third quintile of households in terms of per capita expenditure. The poverty lines were VND 1.16 million per person per year in 1993, VND 1.79 million in 1998 and VND 1.92 million in 2002. In this study, we use the same international poverty line and adjust it for 2004 (VND 2.16 million), based on the annual Consumer Price Index (CPI). We have not used the poverty lines developed by the Ministry of Labour, Invalids and Social Affairs (MOLISA), which reflect regional disparity in rice consumption. 3 Details will be furnished on request. 4 Note that these are group averages and should not be taken to imply that there is no heterogeneity among them. 5 See Pincus and Sender (Citation2008) for a detailed explanation of sampling methods and problems of VLSS and VHLSS. They argue that migrant workers without permanent registration, most of whom were shown to be vulnerably by careful examination of Missing Households and Employment Survey (MHES), were excluded by VLSS or VHLSS. So, some underestimation of vulnerability in our analysis cannot be ruled out. 6 See an excellent summary by Hoddinott and Quisumbing (Citation2003a, b) of methodological issues in measuring vulnerability. They contrast the ex ante VEP measure with ex post measures, such as Vulnerability as Expected low Utility (VEU) proposed by Ligon and Schechter (Citation2003), and Vulnerability as uninsured Exposure to Risk (VER), used by Townsend (Citation1994). We use only the VEP measure because VEU or VER can be constructed only with long panel data set where household response to shocks can be identified. 7 VEP has been widely applied in other studies, such as Tesliuc and Lindert (Citation2002), Sarris and Karfakis (Citation2006), Zhang and Wan (Citation2006), Ding et al. (Citation2007) and Gaiha and Imai (Citation2009). 8 One of the limitations of this definition of vulnerability is that it is sensitive to the choice of z. We have defined the poverty line as the international poverty line defined by GSO, checked the sensitivity of the results by applying different levels of poverty line (i.e. 120% and 80%) and found that the general pattern of the results is similar. Details of the results will be provided on request. 9 For a justification of consumption as a welfare indicator, see Ravallion and van de Walle (2008). 10 We have used White–Huber sandwich estimator to overcome heteroscedasticity in the sample. 11 See the Appendix for definitions of the variables and descriptive statistics. 12 See Chaudhuri et al. (Citation2002), Chaudhuri (Citation2003) and Hoddinott and Quisumbing (Citation2003b) for technical details. 13 See Gaiha et al. (Citation2007) for a further extension of the model. 14 Following Barrientos (Citation2006), we will use chronic poverty and poverty traps interchangeably. 15 A further exposition of the model can be found in Gaiha et al. (Citation2007). 16 On the gender inequality in Vietnam, some recent studies show that female-headed households have lower incidence of poverty (e.g. Asia Development Bank (ADB), Citation2005), which may simply reflect their concentration in urban areas. On the other hand, other studies have found greater income inequality between men and women in recent years arising from disadvantages for the latter in the labour market (ADB, Citation2005; Kabeer et al., Citation2006; Lee, Citation2006). 17 This study uses total land as a general proxy for household wealth. The regression results for rural Vietnam using the disaggregated data of land holding (which are not shown but will be furnished on request) show that aquacultural land is the most significantly associated with higher consumption, followed by agricultural land and forest land. This presumably reflects the increasing role of fishery in the Vietnamese economy, which has emerged as one of the fastest growing sectors during the transition period. According to International Monetary Fund (IMF, Citation2006), average growth rates in agriculture, forestry and fishery in 2001 and 2004 were 3.3%, 0.65% and 8.35%, respectively. 18 The full set of results will be furnished on request. 19 Here again it would be worthwhile to view the differences from a broader perspective that allows for cultural diversity in responses to shocks.