Assessing Asset Indices

Filmer, Deon ; Scott, Kinnon

World Bank, Washington, DC

关键词: ACCOUNTING;    ADULT LITERACY;    ADULT POPULATION;    AFFILIATED ORGANIZATIONS;    ALTERNATIVE ASSET;   

DOI  :  10.1596/1813-9450-4605
RP-ID  :  WPS4605

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

This paper compares how results usingvarious methods to construct asset indices match resultsusing per capita expenditures. The analysis shows thatinferences about inequalities in education, health care use,fertility, child mortality, as well as labor market outcomesare quite robust to the specific economic status measureused. The measures-most significantly per capitaexpenditures versus the class of asset indices-do not,however, yield identical household rankings. Two factorsstand out in predicting the degree of congruence in rankingsbetween per capita expenditures and an asset index.Firstis the extent to which per capita expenditures can beexplained by observed household and communitycharacteristics.In settings with small transitory shocksto expenditure, or with little measurement error inexpenditure, the rankings yielded by the alternativeapproaches are most similar. Second is the extent to whichexpenditures are dominated by individually consumed goodssuch as food.Asset indices are typically derived fromindicators of goods which are effectively public at thehousehold level, while expenditures are often dominated byfood, an almost exclusively private good. In settings whereprivate goods such as food are the main component ofexpenditures, asset indices and per capita consumption yieldthe least similar results, although adjusting for economiesof scale in household expenditures reconciles the results somewhat.