3. GLOBALIZATION AND WORKING CONDITIONS: A FRAMEWORK FOR COUNTRY STUDIES
85
(maintained by the United Nations Statistical Division), the TRAINS database (maintained by UNCTAD), and the World Trade Organization Integrated Database. While the WITS software is free, access to the databases themselves is not. UNCTAD also compiles foreign investment data and maintains two products to make those data available. UNCTAD’s FDI statistics database includes aggregate inflows, outflows, and stocks for 196 countries. Its World Investment Directory contains more detailed statistics on transnational companies in selected countries. Central banks and national government agencies also often collect more detailed foreign investment statistics, but availability of these data varies among countries. Industries
Industry-level data often come from government surveys or mandatory reporting. They are most useful for identifying changes among industries and changes in aggregate variables (such as employment, average wages, or production shares). High-frequency, industry-level data are also very useful for capturing the time paths of different variables, such as relative wages, production shares, and employment, because it is easier to remove the business-cycle component from high-frequency data. An advantage of these data is that they are often easy to find. The main disadvantage is that they lack the level of detail to accurately capture many of the aspects relevant to analyzing the effects of globalization, such as workers’ demographic characteristics. Firms
Firm-level data are ideal for identifying the effects of globalization on working conditions across firms and thereby identifying composition effects. The key, of course, is to have some measure of working conditions at the firm level, which is rare. Firm-level data also allow for the identification of effects of globalization on plant survival (if firms or plants are followed over time). For example, Bernard, Jensen, and Schott (2006) show that competition from low-wage countries increases the probability of closure or contraction of low-productivity plants, but can be positively correlated with expansion of high-productivity plants. The net effect on employment, wages, and working conditions would be revealed, however, at the industry level. The main disadvantage of these data is that they are difficult to find, especially in developing countries. Notable exceptions include Chile (Pavcnik 2002; Alvarez and Robertson 2002), Mexico (Hanson and Harrison 1999; Verhoogen 2007), and Indonesia (Lipsey and Sjoholm 2004a, 2004b). Even when collected, these data may not be available consistently over time, or may be limited to specific sectors (particularly manufacturing). Workers
Worker-level data are found in household surveys, labor force-surveys, and consumption surveys, and are becoming increasingly available for developing countries. Many studies analyze these kinds of data in developing countries. Several studies have used worker-level data to evaluate the effects of globalization on wages, job loss and unemployment duration, and wage inequality. Labor force or household surveys generally include demographic data that enable the researcher to control for education, age, and other worker characteristics that may mask the effects of globalization. Such data also help identify the effects on workers in different demographic groups. Data from household surveys may also include