The relevance of the middle class in social sciences is undoubted; understanding the middle class is fundamental to understand Latin America. Despite the potential importance of the middle class, there is no consensus either on its definition or its measurement. This thesis investigates on the definition of the middle class, in order to provide a more comprehensive understanding of this social group in Mexico, from an income (unidimensional) perspective in Latin American to a multidimensional definition for Mexico. Moreover, this work analyses the Mexican middle-class women’s labour conditions when facing no-fault divorce, and explores the costs on informal loans middle-class individuals face when financially excluded. Most middle-class income definitions (relative, absolute and a proposed bipolarisation analysis) suggest an expansion of the middle class in Brazil, Chile, Mexico, Paraguay, Peru, and Uruguay between 2000 and 2015. Income definitions, nevertheless, do not reflect the multidimensional nature of the middle class. This work proposes a multidimensional definition of the middle class based on counting the middle-class characteristics households have (counting approach), finding that an income definition of $10 and $50 USD per day has a good fit in identifying the proposed multidimensional definition, compared to other income definitions (34.0% identification overlap). Based on a quasi-experimental technique, middle-class women exposed to no-fault divorce participate 3.5 percent points more on the labour market, reduce their participation in the informal sector by 3.1 percent points, and worked on average one hour more than non-exposed middle-class women. Finally, based on a propensity score matching estimation, middle-class individuals in urban areas excluded from formal financial services pay 7.3 percent points more on interest rates for informal loans than non-excluded individuals, despite this loans where granted by relatives, friends and co-workers of the borrower.
