Ortiz-Juarez, EduardoLopez-Calva, Luis F.2012-03-192012-03-192011-12-01https://hdl.handle.net/10986/3669Measurement of the middle class has recently come to the center of policy debate in middle-income countries as they search for the potential engines of growth and good governance. This debate assumes, first, that there is a meaningful definition of class, and second, that thresholds that define relatively homogeneous groups in terms of pre-determined sociological characteristics can be found empirically. This paper aims at proposing a view of the middle class based on vulnerability to poverty. Following this approach the paper exploits panel data to determine the amount of comparable income -- associated with a low probability of falling into poverty -- which could define the lower bound of the middle class. The paper looks at absolute thresholds, challenging the view that people above the poverty line are actually part of the middle class. The estimated lower threshold is used in cross-section surveys to quantify the size and the evolution of middle classes in Chile, Mexico, and Peru over the past two decades. The first relevant feature relates to the fact that the proposed thresholds lie around the 60th percentile of the distribution. The evidence also shows that the middle class has increased significantly in all three countries, suggesting that a higher number of households face lower probabilities of falling into poverty than they did in the 1990s. There is an important group of people, however, which cannot be defined as middle class from this perspective, but are not eligible for poverty programs according to traditional definitions of poverty.CC BY 3.0 IGOABSOLUTE POVERTYABSOLUTE POVERTY LINEABSOLUTE TERMSAGRICULTURAL WORKERSCAPITA EXPENDITURECARIBBEAN REGIONCENTRAL REGIONCROSS-SECTIONAL DATADAILY INCOMEDATA QUALITYDECLINING INEQUALITYDEFINITIONS OF POVERTYDEPENDENT VARIABLEDESCRIPTIVE STATISTICSDEVELOPING COUNTRIESDEVELOPING WORLDECONOMIC DEVELOPMENTECONOMIC DOWNTURNSECONOMIC GROWTHECONOMIC INEQUALITYECONOMIC PERFORMANCEECONOMIC SHOCKECONOMIC SHOCKSECONOMIC TRANSFORMATIONEMPIRICAL WORKFARMERFARMERSGROUP MEANSHOUSEHOLD CONSUMPTIONHOUSEHOLD DATAHOUSEHOLD HEADHOUSEHOLD MEMBERSHOUSEHOLD PER CAPITA INCOMEHOUSEHOLD SIZEHOUSEHOLD STRATEGIESHOUSEHOLD SURVEYHOUSEHOLD SURVEYSHOUSINGHUMAN CAPITALHUMAN CAPITAL INVESTMENTINCIDENCE OF POVERTYINCOMEINCOME DISTRIBUTIONINCOME DISTRIBUTIONSINCOME EQUATIONINCOME GROUPSINCOME LEVELINCOME LEVELSINCOME QUINTILESINCOME RISKINCOME SHAREINCOMESINEQUALITYINSURANCEINSURANCE MARKETSINTERNATIONAL POVERTY LINELABOR MARKETLINEAR MODELLINEAR REGRESSIONLOCAL INSTITUTIONSLOGARITHMIC SCALELONGITUDINAL DATAMEASUREMENT OF POVERTYMEDIAN INCOMEMEDIAN POVERTYMETROPOLITAN REGIONMIDDLE CLASSMIDDLE CLASS CONSENSUSMINIMUM LEVELNEGATIVE SHOCKSPER CAPITA INCOMESPOLICY DISCUSSIONSPOLICY RESEARCHPOLITICAL ECONOMYPOLITICAL SCIENTISTSPOLITICAL STABILITYPOORPOOR HOUSEHOLDSPOOR INDIVIDUALSPOOR POPULATIONPOVERTY ANALYSISPOVERTY DYNAMICSPOVERTY ESTIMATESPOVERTY INCIDENCEPOVERTY LEVELSPOVERTY LINEPOVERTY LINESPOVERTY MEASUREPOVERTY MEASUREMENTPOVERTY PROGRAMSPOVERTY STATUSPRIMARY EDUCATIONPUBLIC INTERVENTIONREGIONAL LEVELSREGRESSION ANALYSISRURALRURAL AREASRURAL RESIDENCESAFETY NETSSANITATIONSAVINGSSCHOOLINGSIGNIFICANCE LEVELSKILLED WORKERSSMALL FARMERSSOCIAL DEVELOPMENTSOCIAL MOBILITYSOCIAL POLICIESSOCIAL PROTECTIONUNEMPLOYMENTURBAN AREASVULNERABILITY TO POVERTYVULNERABLE HOUSEHOLDSWorld Bank10.1596/1813-9450-5902