Statistics in the LAFTA countries are generally in a state of underdevelopment. This is particularly true of labor statistics. The data necessary for the measurement of interindustry, interfirm, and other wage differentials, even in gross terms, are scarce. For intercountry comparisons the problem is more serious because the available data are more heterogeneous.1 Thus, it was essential to conduct a special survey to obtain appropriate data for this study.
The data collection effort was carefully planned. Based on previous experience with similar surveys, the Instituto de Pesquisas Econômicas of the Universidade de São Paulo prepared and tested a preliminary set of procedures for the wage survey. All the ECIEL institutes collaborating on the project attended a workshop to review this proposal. A revised set of survey procedures was adopted at this workshop, which the institutes tested in their respective countries. The participants met again in another workshop, from which a final set of procedures emerged. Most of the surveys were undertaken in 1967, with the questionnaires and manuals used following the basic design approved by the institutes.
The survey work in the field was afflicted by problems in virtually all countries. In some cases, the first attempt was a partial or total loss, because the desired information was either missing or unreliable, and a new or follow-up survey had to be implemented. Long delays resulted from the lackluster collaboration of many firms. All the data refer to the same date ( November 1966), irrespective of the moment at which each survey was undertaken. The period of reference chosen was considered to be as normal as could be expected in a study involving several industries and countries. It should be borne in mind that Latin American manufacturing, and especially its modern sector, has been less affected by cyclical fluctuations than the rest of the economy. In terms of the overall performance of the national____________________