In this appendix, the measurement of wage differentials is carried out using an information theory approach.1 The method is then applied to a sample of firms in Mexico City. As the Mexican survey had less detailed information on workers' characteristics, the application of the methodology used in the other chapters of this book was not feasible.
Two aspects of wage differentials will be considered: vertical occupational disparities basically due to differences in skills possessed by wage earners, and horizontal disparities due to differences among similar occupational categories (jobs or groups of jobs) in different economic units (firms, industries, etc.). To a considerable extent, differentials of the first type reflect the different contributions which employees make to production. Thus, the pay scale is supposedly related to skills, a proxy for intrinsic labor productivity. Within each firm, wages range from the low levels corresponding to unskilled manual workers, to the higher levels paid to technicians and administrative personnel. Ultimately, these differentials are substantially affected by relative scarcities of personnel and the interplay of market forces.
The horizontal differentials result from factors like differences in the cost of living in the areas where the industries or firms are established or in the relative importance assigned to different jobs in each industry or firm. For example, a chemist might not perform the same tasks in the chemical and food industries. Hence, his productivity may vary from one industry to the
NOTE: The author is a research economist at the Centro de Estudios Económicos y Demográficos of El Colegio de México. He thanks Pedro Uribe for valuable suggestions on the use of information theory in measuring differentials. He also wishes to acknowledge the most helpful comments made on various aspects of the chapter by Raúl de la Peña and Daniel Murayama of the Centro de Estudios Económicos y Demográficos of El Colegio de México. He is indebted to the computer centers of the Universidad Nacional Autónoma de México and Petróleos Mexicanos (PEMEX) for data processing and computation.