Structural equation model for the multidimensional measurement of the poverty in Mexico

This paper postulate that the poverty in Mexico can be determined analyzing different hypothetical constructs, just like the official measurements, that carry out several organisms of the Federal Government's, to effect of obtain an unique measurement. It postulate that the measurements of the levels of alimentary poverty, of capacities, of patrimony, the index of social backwardness, that obtains the Social Developmental Secretariat, they can integrate with other socioeconomic measurements such as the Marginalization index, that obtains the National Council of Population. Given the subjectivity of the different hypothetical constructs, as poverty, social backwardness and marginalization, the only form of measuring them is indirectly, by using methods that use variables that they can be measured directly. For the integration of the diverse official measurements proposes a method of multivariate analysis known as a Structural Equation Model. It develops a complete model of structural regression, compound by two sub models: the Measurement and the Structure models; which used i variables observed to define j variables latent hypothetical, both dependent and independent, to determinate an integrated measure of the poverty.