During the 1980s, many economic historians studying Latin America focused on the impact of the Great Depression of the 1930s. Most of these historians argued that although the Depression began earlier in Latin America than in the United States, it was less severe in Latin America and did not significantly impede industrial growth there. The historians' argument was grounded in national government records concerning tax revenues and exports and in government-sponsored industrial censuses, from which historians have drawn conclusions about total manufacturing output and profit levels across Latin America. However, economic statistics published by Latin American governments in the early twentieth century are neither reliable nor consistent; this is especially true of manufacturing data, which were gathered from factory owners for taxation purposes and which therefore may well be distorted. Moreover, one cannot assume a direct correlation between the output level and the profit level of a given industry as these variables often move in opposite directions. Finally, national and regional economies are composed of individual firms and industries, and relying on general, sweeping economic indicators may mask substantial variations among these different enterprises. For example, recent analyses of previously unexamined data on textile manufacturing in Brazil and Mexico suggest that the Great Depression had a more severe impact on this Latin American industry than scholars had recognized.