5.1. Poverty Indicators
Poverty is a multi-faceted phenomenon which affects not onlythe ability to purchase goods, but also vulnerability towards various pressuresthat may prohibit an individual from enjoying life. This vulnerability may begauged from living conditions such as employment, health, education, andhousing. It is important to monitor gender differences in poverty, vulnerabilityand living conditions, and also to understand the causes of these differences,in order to prepare strategies for more efficient intervention schemes aimed atpoverty reduction.
Poverty typically is measured by purchasing power or percapita expenditures made by the household, in the form of poverty rates orexpenditure quintiles. Purchasing power has a strong correlation to most otherliving condition indices and is therefore used as a main indicator of povertyand vulnerability. Productivity and incomes from occupations and livelihoods areimportant factors for reducing poverty. Social conditions such as health,nutrition, education and housing influence productivity, thus affecting povertystatus. These in turn are influenced by poverty, affecting the ability ofhouseholds to gain access to adequate social conditions to improve theirproductivity. Efforts towards poverty alleviation therefore, require a completeintervention scheme, not simply in economic aspects, but including socialdimensions as well, so that poverty may be addressed as a socio-economicphenomenon.
5.2. Variables in Poverty and LivingConditions
Analysis of determinants of poverty is essential forpreparing strategies towards efficient intervention.
This section of the report presents results of a multi-variateanalysis of the relationship between various social and economic aspects ofliving conditions and poverty as measured by purchasing power. The data providesa picture of the living conditions in Viet Nam, and evaluates the variousprobable determinants of these living conditions. It provides valuableindicators of poverty in order to inform suggestions for poverty reductionpolicy reform in the light of gender analysis.
Multi-variate analysis of the likely determinants of povertywas conducted separately for female-headed households and all households (amajority of which are male-headed) and for male and female individuals. Theresults show the probable differences in factors such as education and place ofresidence that affect poverty amongst FHHs compared to all households, whenother characteristics such as education, age, ethnicity, etc. are the same.Poverty is typically determined at a household level. Therefore, this sectionfocuses primarily on the regression results of FHHs versus all households ratherthan individual level results.
Interactions of economic and social dimensions ofpoverty must be researched, as strong inter-linkages exist.
The analysis is presented in terms of location of householdsin order to identify geographic determinants of poverty when other factors suchas educational level, ethnicity and employment are held constant. Such analysiswill help determine whether geographic targeting (with lower administrativecosts) or other forms of targeting form more appropriate poverty reductionstrategies.
The link between poverty and rural residence is strongbut appears to be more important for FHHs than for allhouseholds.
Rural residence is strongly correlated with poverty overall,but more so for FHHs compared to all households. For individuals, when the sexof the household head is held constant, rural residence has a higher impact onthe probability of an individual living in poverty amongst males than amongstfemales. Clearly the relationship between gender, urban/rural residence andpoverty is complex. Nevertheless, targeting of poverty reduction efforts forboth men and women in rural areas is important, and special efforts aimed atFHHs may be required.
Regional and provincial differences in probability ofbeing in poverty among FHHs suggest that geographic targeting may beimportant.
Region of residence also has a strong association withhousehold poverty. For all households, when other factors were held constant,residence in provinces of the Southeast region led to a lower likelihood ofpoverty compared to other regions. Other regions had the same or higherprobability of poverty compared to Bac Lieu province in the Mekong Delta whichwas the comparator. However, FHHs in most provinces exhibited higher probabilityof poverty than in Bac Lieu province, which is a relatively poor province.Further research may be required into these provincial level effects beforegeographic targeting is used for gender-specific poverty reductionprogrammes.
Targeting of gender-specific interventions among ethnicminorities will be important.
Female headed households from Kinh and Muong groups had lowerlikelihood of poverty compared to other ethnic groups. This finding supportsearlier studies showing a need to focus gender based poverty interventions amongethnic groups other than among the Kinh or ethnic groups that closely resemblethe Kinh.
Educational improvements pay off forFHHs.
The higher the educational attainment, the lower thelikelihood of poverty for all households and for FHHs, even with occupation andgeographic residence held constant. The greater the share of household membersreceiving apprenticeships or informal training, the lower the odds of povertyoverall, but for FHHs, the impact is slightly stronger.
A clearer definition of what constitutes a FHH householdis warranted both for research and targeting purposes.
The data analysis does not show an association between age ofhousehold head and poverty within FHHs. However, in the analysis of allhouseholds, older household heads still of working age tend to have a lowerlikelihood of poverty. As a household head grows older, experience, accumulatedcapital and greater labour supply (due to less childcare, older aged children),is typically associated with lower poverty. However, the nominated householdheads in FHHs may not be the true decision-making household head, but rathersomeone, usually older, selected for administrative reasons. The relationshipbetween age of household head and poverty may not be so clear, and inferencesshould be used with caution.
Single person households may require specificinterventions.
Household size does not affect the probability of FHHs inpoverty, except for the case of one-person households where the likelihood ofpoverty for one-person FHHs is substantially higher than for MHHs. Single-personhouseholds typically involve a lack of labour, which is more detrimental towomen than to men when other factors are held constant. The group ofsingle-person households is likely to be small, and overall not very likely tobe in poverty, but single female households should nevertheless be considered asa special target group for interventions. More detailed analysis of laboursupply within different types of households may be required to understand genderdifferences in poverty.
