A Multilevel Analysis of Poverty among Mexican Americans and Immigrants in the Southwest United States

Ginny Garcia, University of Texas at San Antonio

Data from the 2006 American Community Survey are used in conjunction with 2000 Decennial Census data in an effort to understand the effects of individual and contextual level characteristics on the prediction of poverty for Mexican Americans and Mexican immigrants in the Southwest United States. Such variables as immigration status, employment status, and number of children, among others, are utilized at the individual level in order to predict the likelihood of being in extreme poverty, one hundred percent poverty, and low-income. Data are then introduced at the contextual level (level-2) (Super-PUMAs containing at least 400,000 persons) measuring percentage of persons in poverty, percentage of Mexican Americans and Hispanic immigrants in the PUMA, and presence of particular occupational industries. Many studies have focused on the individual level predictors of poverty; this research goes one step further and predicts poverty not only with respect to individual predictors but also group level variables.